diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fb169f969ac5a922f6085d5c0d761080ef7cd9fe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e287555f09fc8fbd844ed506cef82594fa9e05d6f8d1580932eccce83d6f5cb3
+size 148732
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..a60f2619c7ca28135d4877cb87849e2043bb62f1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen in a kitchen, holding a piece of food in his hands. He is wearing a black t-shirt with a graphic design on it and has a tattoo on his arm. The kitchen is equipped with a refrigerator and a microwave, and there is a sink visible in the background. The man appears to be in the middle of eating, as he is holding the food up to his mouth. The overall style of the video is casual and candid, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "EyJYlzmV3eM_17_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a piece of food, a black t-shirt with a graphic design, a tattoo on the arm? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting at a table, holding a piece of food. He is wearing a black t-shirt with a graphic design, and a tattoo is visible on his arm. All specified objects are clearly present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in in a kitchen, holding a piece of food, He is wearing a black t-shirt with a graphic design and and there tattoo is visible on his arm. The the elements are present present in match the description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b17f0835c8f5c98f68f367eeedb28f0a59e6e0db
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0ad42d0d313e708cc6f85d4463bc5cca4876937098513b3a3aa0226477882652
+size 176134
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..415a66e0c6f2fcee5ae12ecbb2a73346e8806cb1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with glasses sitting on a purple couch. He is wearing a gray t-shirt and appears to be in a relaxed setting. The man is looking directly at the camera, suggesting that he is engaged in a conversation or an interview. The background is simple, with a white wall and a framed picture hanging on it. The overall style of the video is casual and informal, with a focus on the man and his surroundings."
+    ],
+    "video_ids": [
+        "1GJyOY9whPk_9_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with glasses, wearing a gray t-shirt, sitting on a purple couch and looking directly at the camera.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing glasses and a gray t-shirt, sitting on a purple couch, and looking directly at the camera. These elements match the core description provided. Additional elements, such as a framed picture on the wall and a light switch, do not contradict the description and are acceptable as background details.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing glasses and a gray t-shirt, sitting on a purple couch. and looking directly at the camera. These elements match the description description provided. There elements such such as the microphone picture on the wall and a colorful blue, are not contradict the main and are acceptable.\"\n long details.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..de1d05f27d232723c95ef6a62712d368cad21cda
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e91476451bde91d8c90b61cba3a8d1583449e82539eea135a4cd589e55de0563
+size 145303
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..72e594e8937dc923f2bbf43ad947c633ef6bdd95
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of camaraderie between two football players from opposing teams. The player in the purple jersey, representing the Minnesota Vikings, is seen smiling and engaging in a friendly conversation with the player in the white jersey, representing the Chicago Bears. They are standing on the field, surrounded by other players and staff, with the stadium in the background. The atmosphere is one of sportsmanship and mutual respect, as they share a light-hearted moment before the game begins."
+    ],
+    "video_ids": [
+        "YXzABP5A4IE_5_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two football players, one in a purple jersey (Minnesota Vikings) and one in a white jersey (Chicago Bears).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two football players, one wearing a purple jersey with 'Vikings' and the number 8 (Minnesota Vikings) and the other wearing a white jersey with the number 10 and the Chicago Bears logo (Chicago Bears). The jerseys, team names, and numbers are clearly visible and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two football players, one wearing a purple jersey with theMinnesotaikings' written the number '32which Vikings) and the other wearing a white jersey with ' number 86 ( a logo Bears logo.Chicago Bears). The players and helmets names, and numbers match consistent visible, match the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2efe3c15a5b11f3d21c3e488359d9a680b9ba56f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03f3b90ecac7fcb14637fa6ec5c344247e141d686a053938a56d8fab09ac1b94
+size 260516
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..3da2809ea4e2b1b26d4c2937b36529073014dd20
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a blue Tesla Model X driving on a muddy road. The car is sleek and modern, with a distinctive design that includes a large rear window and a unique rear light. The car's black wheels are covered in mud, indicating that it has been driving on the dirt road. The road itself is muddy and appears to be unpaved, with trees and bushes lining the sides. The car is moving forward, and the video captures the motion of the car as it travels down the road. The overall style of the video is realistic and naturalistic, with a focus on the car and its surroundings."
+    ],
+    "video_ids": [
+        "70b-VHqd9tw_50_25to227"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue Tesla Model X? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a blue Tesla Model X parked on a muddy road, which matches the description. The car's distinctive design, including its shape, windows, and wheels, is consistent with a Tesla Model X. The color is a shade of blue, and there are no elements contradicting the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a blue vehicle Model X driving on a rocky,, which align the description of The car's design shape, including its sleek and color, and wheels, is consistent with the Tesla Model X. The environment of also vibrant of blue, and the are no additional ining the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..922c8568064b4dfa6652ca3e24301906d8148dfd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dcad7383b0b03c1ef22051c23af33969794083753cd8da5f208cfe3387a1f9dd
+size 439718
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..5b362c1bd51fca59a57ddf52c09910d1b4c45509
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a short, visually appealing sequence of three frames, each depicting a different stage of a healthy breakfast. The style is clean and modern, with a focus on vibrant colors and fresh ingredients.  In the first frame, a white bowl filled with a variety of fruits and granola is placed on a wooden table. The bowl contains a mix of berries, grapes, and orange slices, along with a scoop of granola.  In the second frame, a wooden spoon is inserted into the bowl, ready to stir the ingredients together. The spoon's handle extends out of the bowl, inviting the viewer to take a bite.  In the third frame, a glass jar of milk is placed next to the bowl. The milk is poured into the bowl, creating a creamy and nutritious base for the fruit and granola. The wooden table provides a warm and rustic backdrop to the entire scene.  Throughout the video, the focus remains on the bowl and the ingredients, with the wooden table and spoon serving as supporting elements to enhance the overall aesthetic. The video does not contain any text or additional elements, allowing the viewer to fully appreciate the simplicity and beauty of the healthy breakfast."
+    ],
+    "video_ids": [
+        "9tOgs41ZkvY_65_22to243"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white bowl, a wooden spoon, and a glass jar of milk? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a white bowl filled with various fruits and granola, a wooden spoon resting on the rim of a glass jar containing milk, and the jar itself. These core objects are present and accurately depicted, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a white bowl filled with various fruits and nutsola. a wooden spoon being in the bowl of the bowl jar, milk, and a pouring of. The elements elements match present and match depicted, fulfilling the 'Object(s)' condition as}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..add93069746947fc490068d2975ab5b7af2645f3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f177f0986e0f932dc36102606a574e1a8e1a2fab05190e1a82cef98e1c4e5997
+size 131986
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..459277fd7fb9e93442a7ae609fa4bf5712e21b71
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in front of a car, speaking into a microphone. He is dressed in a suit and tie, and appears to be presenting or announcing something. The car behind him is sleek and modern, with a prominent logo on the side. The setting suggests a formal event or press conference, possibly related to the car or the company it represents. The man's demeanor is professional and confident, indicating that he is an authority figure or spokesperson for the event. The overall style of the video is polished and professional, with a focus on the man and the car."
+    ],
+    "video_ids": [
+        "8vNPJxasV14_34_30to154"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a suit speaking with hand gestures, and a car is visible in the background. These elements match the 'Object(s)' condition described, with no contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man in a suit standing at a gestures, which there car is visible in the background. The elements match the 'Object(s)' condition described, with the additional or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a602bae7df1175df69394c84d8569b1e0891c88e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1118-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ee94924b8c17f9c83fa2b8b6ca09ada0b7582acc0f479a0199216231dc3fb37
+size 100639
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..59f3fa0bf326d44ab72c160b2d8d24d9a3afc6b7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "Three young men are sitting on a blue couch in a colorful room with a brick wall and geometric shapes. They are wearing casual clothing and appear to be engaged in a conversation. The room has a modern and vibrant atmosphere."
+    ],
+    "video_ids": [
+        "Y6aZR4xHzFY_40_20to224"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three young men sitting on a blue couch, wearing casual clothing and engaged in a conversation.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three young men sitting on blue couches, dressed in casual clothing, and appearing to be engaged in a conversation. Their body language and positioning suggest active participation in a discussion, and the setting supports a relaxed, conversational atmosphere. The presence of a TV screen and decorative elements does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three young men sitting on a couches, wearing in casual clothing, and engaged to be engaged in a conversation. The attire language and the suggest they interaction in a discussion, which the setting is the casual, informalational atmosphere. The description of additional colorful and in a elements in not contradict the core description but}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6e6c5feee22a66bc2c640cd4765c349be80694aa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a2143aaba3b0fa3cbda39e06aed244d12d2571d7a9ebfa00e04363a613601f1
+size 128331
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..5d2300b9d403cce4955abbebfef801060f961ce3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a scene from a video game set in a village nestled in a mountainous area. The village is constructed of wooden buildings with stone walls, and colorful flags hang from the rooftops. The village is surrounded by lush greenery and towering mountains, creating a serene and picturesque setting. The village is bustling with activity, with villagers going about their daily routines. The video is rendered in a realistic style, with attention to detail in the textures and lighting. The overall atmosphere of the video is one of tranquility and harmony with nature."
+    ],
+    "video_ids": [
+        "-qD-wHvnOqU_32_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Wooden buildings, stone walls, colorful flags, lush greenery, towering mountains, villagers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing wooden buildings with stone walls, colorful flags strung across the scene, lush greenery including trees and foliage, and towering mountains in the background. While no villagers are visibly present, the setting strongly suggests a village environment, and the absence of visible villagers does not contradict the core description, as the condition does not require explicit depiction of villagers, only the presence of the other listed elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting wooden buildings with traditional walls, colorful flags,ung between the street, lush greenery, palm and plants, and towering mountains in the background. The there villagers are explicitly present in the overall and suggests a village environment, which the presence of villagers villagers does not contradict the core description. as the focus only not specify their depiction of villagers.\"\n only the implication of the described elements elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f12552998d86ee155d2b50fce1f4c53f4dfa1258
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:062ef64ca0e739001378cc397a078a445c1935b984ead1a2053e362a0b471874
+size 226817
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..4adc27180ba2ac05f7f4010989c93ab903dacbd4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features three animated buses, each with a unique color and design, driving down a city street. The first bus is red, the second is yellow, and the third is blue. Each bus has a face drawn on its front, giving them a friendly appearance. The street they are driving on is lined with buildings and trees, creating a typical urban environment. The buses are moving in the same direction, suggesting they are part of a convoy or a fleet. The overall style of the video is cartoonish and playful, with a focus on the buses and their journey through the city."
+    ],
+    "video_ids": [
+        "HQjt48_gBSU_58_0to154"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three animated buses, each with a unique color and design: red, yellow, and blue.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows three animated buses: one red (number 1339), one yellow (number 02), and one blue (number 120). Each bus has a distinct color and design, matching the description. The presence of a parking lot, bushes, and other background elements does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows shows three animated buses, a red,withed1),),),), one yellow (number 102 and one blue (number 033). Each bus has a unique color and design, matching the description provided The presence of additional cartoon meter and trees, and buildings elements elements does not contradict the core description of}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..768b62bbb7efb81e9c8a480d4a86a9dbed99ebdc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c0baef0730658768863d1c91af7091053bf6df44d8b2406ced7ebaa653cf891
+size 181164
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..7775b5fc4deb8086d9f5106a53ef5396d1372889
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features an older man with white hair and glasses, wearing a suit and tie. He is standing in front of a wall with a poster of three men in black shirts. The man appears to be speaking, as his mouth is open. The style of the video is a straightforward interview or presentation, with the man as the main subject. The background is simple and uncluttered, focusing attention on the man and his message. The lighting is bright, suggesting an indoor setting. The overall tone of the video is professional and serious."
+    ],
+    "video_ids": [
+        "I_oFRIQwtgg_25_0to143"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An older man with white hair and glasses, wearing a suit and tie; a poster of three men in black shirts.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an older man with white hair and glasses, wearing a suit and tie, which matches the description. In the background, there is a poster featuring three men in black shirts, also matching the description. There are no significant contradictions with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an older man with white hair and glasses, wearing a suit and tie, which matches the description. Additionally the background, there is a poster of three men in black shirts, which align the description. The are no contradictions contradictions or the given description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0e6e88e9e3dafb4756606c60094408c9f9331980
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4f3dcacb12378e4a873359dac2f55851d664dc1745061c64d240d8a7d2df4486
+size 202904
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..b10f55d2f227a957c2f76f72b070ddb1662f7624
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a professional interview with a man in a white shirt and blue tie. He is seated in a chair in front of a white wall. On the wall, there are three photos of men in suits. The man in the white shirt is looking directly at the camera, and his hands are clasped together. The style of the video is formal and professional, with a focus on the man being interviewed. The background is simple and uncluttered, allowing the viewer to focus on the man and his responses. The lighting is bright and even, highlighting the man's features and the details of the photos on the wall. The overall impression is one of seriousness and professionalism."
+    ],
+    "video_ids": [
+        "EbT1InAXX6Q_9_0to126"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a white shirt and blue tie, three photos of men in suits.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man wearing a white shirt and blue tie, as described. Behind him, there are three visible photos of men in suits (or formal attire), which aligns with the description. The presence of additional elements, such as the partially visible person in the foreground and the logo in the corner, does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a white shirt and a tie, which described. Additionally him, there are three photos photos of men in suits,two at attire), which alsos with the second. The presence of the elements ( such as the man visible photo in the foreground, the office in the background, does not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4d22d7b23d214feb9ecb139c295dd2b6670805b7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:787919c43b76e25bebca5fd4e9201c102dd870b2c3aeb64362eaebed8e5b71f4
+size 53820
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..3fbd57ee3425a71aaf1a873618e219d36b8e9608
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is an aerial view of a city skyline at sunset. The city is densely populated with tall buildings of various sizes and designs, reflecting the warm hues of the setting sun. The buildings are clustered together, creating a sense of urban density. The city is surrounded by a large body of water, which adds a serene element to the bustling cityscape. The sky is a gradient of colors, transitioning from the warm oranges and yellows of the sunset to the cooler blues of the evening sky. The overall style of the video is a blend of urban architecture and natural beauty, capturing the essence of a city at the end of a day."
+    ],
+    "video_ids": [
+        "5cZGlIi1myA_18_0to103"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Tall buildings, large body of water? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows numerous tall buildings in a dense urban area, and a large body of water is prominently visible adjacent to the cityscape, fulfilling the specified conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a tall buildings densely the dense urban area, which there large body of water is visible visible in to the cityscape. fulfilling the ' conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f575a6f2584f91453c652ddeb9b2de4a8f0e010d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8638fe1d02397c51636a3e48eddf1c353b72371d2de238dcd4a1c7129446e334
+size 168484
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..0af2e96aa9c1e44ba7a29d58672cfaf20ce36fbb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a glass jar with a metal lid, filled with white powder. The jar is placed on a wooden surface. In the first frame, the jar is closed, and the lid is securely fastened. In the second frame, the jar is tilted slightly, and the powder begins to spill out. In the third frame, the jar is tilted more, and a significant amount of powder has spilled out onto the wooden surface. The style of the video is a simple, straightforward time-lapse, capturing the gradual process of the powder spilling out of the jar. The focus is on the jar and the powder, with no additional elements or distractions. The video does not contain any text or sound."
+    ],
+    "video_ids": [
+        "6kKiQ-nUp5A_9_0to138"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A glass jar with a metal lid, white powder, wooden surface? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video includes a clear glass jar with a metal clamp lid, which is lying on its side with white powder spilling out onto a wooden surface. These elements are clearly visible and match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a glass depiction jar with a metal lid lid, containing is filled on a side on white powder insideilling out. a wooden surface. The elements match consistent visible and match the description provided.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..acceabbaacb414cb4dd6ada6dffa426a07cf6da1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d8a1bb1c664d503771f6ef88163c582f81f5df5eafd01b796ac9dbb820c16a22
+size 180975
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..1f33e1c6f8d4ec2b57e6cb25cdf192349af7fb0e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a bustling city scene with a multi-lane highway filled with a variety of vehicles, including cars, buses, and trucks. The highway is surrounded by lush greenery and trees, providing a stark contrast to the urban environment. The vehicles are moving in both directions, indicating a busy time of day. The highway is lined with multiple traffic signs, including speed limit signs and directional signs, guiding the flow of traffic. The overall style of the video is a dynamic and colorful representation of urban life, capturing the energy and movement of the city."
+    ],
+    "video_ids": [
+        "32t8d2eQJAA_67_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Vehicles (cars, buses, and trucks), traffic signs (speed limit and directional).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows numerous vehicles including cars, buses, and trucks moving along a multi-lane highway. Traffic signs indicating speed limits (e.g., 60, 70, 80 km/h) and directional guidance (e.g., '\u5c0f\u578b\u8f66' for small vehicles, '\u5927\u578b\u8f66' for large vehicles) are prominently displayed. These elements fully satisfy the 'Object(s)' condition as described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a vehicles, cars, buses, and trucks, along a busy-lane road. There signs are speed limits ande.g., 60 km 80) 80)/h) and directional information aree.g., 'K\u8f66' which small vehicles) '\u5927\u578b\u8f66' for large vehicles) are also displayed. The elements align meet the 'Object(s)' condition as described.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d574ff61732b0cbe5aaab0cada395398e734f675
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1119-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0e02d307b67e038c6550b86bd8f733e52b0dea96f243d74b251792edeba55a00
+size 282014
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..4bdf472985e7de7a391f928e39e36709fbe0d2e4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a kitchen, preparing a meal. She is wearing a red apron and a white shirt. The kitchen is well-equipped with various appliances and utensils. In the first frame, she is seen holding a bowl, which she places on the counter. In the second frame, she is seen pouring something from a bottle into the bowl. In the third frame, she is seen holding a knife, presumably to cut something. The kitchen has a microwave, an oven, and a sink. There are also several bottles and a vase on the counter. The woman appears to be in the middle of cooking, and the video captures her actions in a realistic and detailed manner."
+    ],
+    "video_ids": [
+        "0Qp3Ms6IWos_5_44to205"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a bowl, a bottle, a knife? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a kitchen setting, wearing an apron and standing near a counter with various items. A black cast iron skillet (which can be considered a type of bowl) is visible on the counter, along with a green bottle (likely dish soap or hand soap) and a knife block containing multiple knives. All specified objects are present in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a kitchen setting. which a apron, engaging at a countert. a kitchen. She bowl bowl- skillet iswhich can be considered a bowl of bowl) is on on the counter. and with a bottle bottle andpossibly a soap) a sanitizer) and a knife.. a knives. The the objects are present and the scene,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..96e800a7382486149786fb6d8fc798f9f07419ec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26b51fc4f37256e793496a09040148662307e74465eb2fa8b886dbe132811fd5
+size 175731
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..b7124b24784f1f50173777a792796a0b88302d4a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment from Fox News, featuring a man in a suit and pink tie. The man is speaking on a topic related to Oprah Winfrey campaigning for Georgia's democratic candidate for governor. The style of the video is a typical news broadcast, with a professional and serious tone. The man is the main subject of the video, and he is positioned in the center of the frame. The background is a simple, uncluttered news studio, with a blue and white color scheme. The video is likely to be informative and informative, providing viewers with news and updates on the political campaign."
+    ],
+    "video_ids": [
+        "1lf4iTaykgo_29_166to376"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and pink tie.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed in a suit and wearing a pink tie, which matches the core description. The presence of news graphics and the setting does not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing in a suit and wearing a pink tie, which matches the description description provided The background of text-related and text text of not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3348ee3b61d8efe425e36ac0dae470723438f19c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dee382f2e949d06e43bc2fdbbd1cbc5fbd6481a794856d1a31d6985ee8585a63
+size 155631
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..a1765b820bf6cb16af15ebd5133abbccb574598e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person in a white glove meticulously decorating a series of green macaroon cookies with yellow cream and red accents. The cookies are arranged on a metal tray, and the person is using a small tool to apply the cream and accents. The style of the video is a close-up, time-lapse shot that captures the intricate details of the cookie decoration process. The focus is on the hands and the cookies, with the background being out of focus. The video is likely intended to showcase the artistry and precision involved in cookie decoration."
+    ],
+    "video_ids": [
+        "5TJKJ2d5U8g_94_424to583"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person, green macaroon cookies, yellow cream, red accents, metal tray, small tool? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person wearing gloves assembling green macaroon cookies with yellow cream and red accents on a metal tray. The cookies are shaped like cute animals, and the person uses their hands as the tool to assemble them. All core elements from the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person decorating a decorating green macaroon cookies with yellow cream and red accents. a metal tray. The small are being like Christmas faces, and the person is a hands and a small to place them. The the elements ( the description are present in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1464be285dcf526e5a74631a1be45a802c33500a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c05864ff4d4dc795c319e0c0c589c1d16cf0d928c097cba8ffeaedaad015ee93
+size 131800
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..2e4bc7af19746e7d4d004a3daad466e521f71bc3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen sitting at a table with a variety of hot sauce bottles in front of him. He is wearing a gray hoodie and appears to be in a dark room with a black background. The man is holding a bottle of hot sauce in his hand, and his facial expression suggests that he is about to taste it. The hot sauce bottles are of different sizes and colors, indicating a variety of flavors. The man's casual attire and the informal setting suggest that this could be a casual or informal event, possibly a hot sauce tasting or a challenge. The focus of the video is on the man and his interaction with the hot sauce, with the background serving to highlight the subject."
+    ],
+    "video_ids": [
+        "0-BQ_yTNyTs_128_163to296"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a table, and several hot sauce bottles.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man seated at a table with multiple hot sauce bottles arranged in front of him. The core elements described \u2014 a man, a table, and several hot sauce bottles \u2014 are all present and accurately represented. Additional elements (like a pitcher and logo) do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man sitting at a table with several bottles sauce bottles in in front of him. The man elements of in a man, a table, and several hot sauce bottles \u2014 are all present and accurately depicted in The elements,like the background of a in do not contradict the core and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a3ab7281e302e16d28c8707b628df4c3879cd3a5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57321f02cc78c7bd8ceffa9776269fc465a202c85d439ccb4f2bb61f07fc915b
+size 137970
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..3cb86762df4e9a683e26097ec741c5b4f5fabece
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the delicate beauty of a purple flower in bloom. The flower, with its vibrant purple petals and yellow center, is the main focus of the video. It is set against a backdrop of green foliage, which provides a natural and contrasting color to the flower. The video is shot in a close-up style, allowing the viewer to appreciate the intricate details of the flower. The flower is shown in different stages of bloom, from the initial opening of the petals to the fully bloomed flower. The video is a celebration of nature's beauty and the simple joy of observing a flower in bloom."
+    ],
+    "video_ids": [
+        "x-wBJnxy4Dg_52_48to175"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A purple flower with vibrant purple petals and a yellow center.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two purple flowers with vibrant purple petals and a yellow center, matching the description. The background and additional stems do not contradict the core description of the flowers.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a purple flowers with vibrant purple petals and a yellow center, which the description provided The flowers is additional elements are not contradict the core description.\"\n the flowers.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..967f9e7cdec57622f3868ebacb6cbc3f18585f08
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9504521ee7b2665838275177964dcfb25dd7e4f6100af71671690d5e4c260037
+size 141629
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..63dcad8be280d3c3297be2d7e035b546b23c6918
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen preparing a cake batter using a red mixer. The person is wearing a green apron and is using a wooden spoon to mix the batter in a glass bowl. The mixer is placed on a kitchen counter, and there is a spatula nearby. The person is carefully stirring the batter, ensuring that it is well mixed. The scene is set in a kitchen, and the person appears to be focused on their task. The video captures the process of preparing a cake batter, from mixing the ingredients to using the mixer to blend them together."
+    ],
+    "video_ids": [
+        "VEQ-EQcDejA_25_28to221"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a red mixer, a wooden spoon, a glass bowl, and a spatula.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person wearing a green apron, using a red stand mixer with a glass bowl attached. The person is also using a wooden spoon and a spatula to scrape down the sides of the bowl. All the specified objects are present and correctly identified in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a person using a green apron, a a red mixer mixer to a white bowl attached. A person is also using a wooden spoon to a spatula to mix the the sides of the bowl. The the objects objects are present and used used in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f4422aa7321753829aa8d68867d6d3ca9df6db23
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:212359a6dbb993996af4dd0a762fcf7bee86a86084745105f45eaa29ec2a5c7e
+size 232904
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..ea62eb65b6143316391aa0dd995baac1bae4ee6d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young man with a beard and a colorful shirt is the main focus. He is standing in a room with a window, and he is looking down at something in his hands. Behind him, a woman with blonde hair and a blue shirt is standing, and she is looking at him with a concerned expression. To the right of the woman, a man with sunglasses and a black shirt is standing, and he is also looking at the young man. The room has a white wall and a window with blinds. The video captures a moment of interaction between the three people, with the young man seemingly showing something to the woman and the man."
+    ],
+    "video_ids": [
+        "YkcvqngBSDI_57_626to780"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man with a beard and a colorful shirt, a woman with blonde hair and a blue shirt, a man with sunglasses and a black shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young man with a beard and a colorful shirt, a woman with blonde hair and a blue shirt, and a man with sunglasses and a black shirt, matching the described subjects. The presence of additional elements does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man with a beard wearing a colorful shirt, a woman with blonde hair and a blue shirt, and a man with sunglasses and a black shirt. all the description elements. The presence of these elements, not conflict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..799b0e2d691250741f48ae40b3d77820a1e5c57f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8bacfbc32fdd3354bfe63f0028f7088f81e81a9168e5b41d65225d2923f9039a
+size 252659
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..81fc14c0ffce726bf4e07ed12e6f670e2778c095
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a blue shirt speaking. He is indoors, with a blue wall in the background. There is a white air conditioning unit to his left, and a shelf with various items, including a calendar and a picture frame. The man appears to be in a room with a casual setting."
+    ],
+    "video_ids": [
+        "NxRNf6tkHVw_54_0to133"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a blue shirt, white air conditioning unit, shelf with various items.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a blue shirt, a white air conditioning unit in the background, and a shelf with various items such as framed pictures and a calendar. These elements align with the specified 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing a blue shirt, standing white air conditioning unit mounted the background, and a shelf with various items. as books pictures and what green. The elements match with the description 'Object(s)' condition, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..052e2bfa2d28bccba987f53220302a425fcdd87e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-112-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b706a790c8733bab0fab16c42751745857a2799e13cc433a2f322a73743d900
+size 133247
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..3639bc40e3a7508ed05ad89c9f8eaea073d1723c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in front of a building with large windows. He is wearing glasses and a suit with a striped shirt. The man appears to be speaking or listening intently. The style of the video is a straightforward, candid shot with no additional effects or filters. The focus is on the man and his interaction with the environment around him."
+    ],
+    "video_ids": [
+        "Ular61rpIVs_181_0to138"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a building with large windows.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in professional attire standing outdoors. Behind him, there is a modern building with large windows, which matches the specified condition. The presence of trees and sunlight does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video features depicts a man wearing a attire, in, The him, there is a building building with large windows, which align the description '. The presence of the and a further not contradict the description description but}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7c77c47f3d7f67f1baacb32fd240ff2b9b61f7e0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c45689657e705c38763e5985fd0dc4dc928d4ee003f8e47373a5f44e0afc3c86
+size 246789
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..bb927aab51fb0e8d2f70eebf606f297e8f46c91b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a vibrant green Audi car parked on a gravel driveway. The car is positioned at a slight angle to the camera, allowing a clear view of its rear and side. The car's design is sleek and modern, with a distinctive black roof and black rims that contrast with the bright green body. The car's license plate reads \"KN69 LYS\". In the background, there's a house with a brick facade and a chimney, as well as a tree with bare branches. The sky is overcast, suggesting a cool or cloudy day. The overall style of the video is realistic and straightforward, focusing on the car and its immediate surroundings without any additional embellishments or distractions."
+    ],
+    "video_ids": [
+        "dWvWR2KJzjs_26_55to272"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A vibrant green Audi car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a vibrant green Audi car, specifically an Audi RS Q3, from a rear three-quarter angle. The car's color, model identification, and brand logo are all visible and match the description. Additional background elements like trees and a house do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a vibrant green Audi car, which an Audi RS43, which the rear three-quarter view. The car's color and model,, and the logo are consistent consistent, consistent the description. The elements elements like the and a building do not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e8a2e9b5d825e403692f600714c404900aafbe85
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:618c3e728e179e85b88dadad7e3d142dcd1932358bc17d408894544c27f33ab3
+size 340059
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..77b7af92111dcdab744dcfd8423d3e38420e410b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows an elderly woman with white hair, wearing a black tank top and a necklace, sitting at a table. She is looking down at a glass of water in front of her. The setting appears to be an outdoor patio or garden area, with wooden chairs and a table visible in the background. The lighting suggests it is daytime. The style of the video is candid and natural, capturing a quiet moment in the woman's day."
+    ],
+    "video_ids": [
+        "TcNtdHOPfMM_28_0to181"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Elderly woman with white hair, wearing a black tank top and a necklace, sitting at a table with a glass of water in front of her.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an elderly woman with white hair, wearing a black tank top and a necklace, sitting at a table. She is holding a glass of water and drinking from it, which matches the core description. Additional elements like other people and background furniture do not contradict the main subject's depiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an elderly woman with white hair, wearing a black tank top and a necklace, sitting at a table. In is holding a glass of water, there from it. which align the description description provided The elements like the glasses or a details are not contradict the main description and description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..aaf7d03af58a08638193ee1f0a3eb64350958253
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d2eea9d7a431a9a0bd796da4aff7ccf2ca695ca578cbea78022baee86c05fc39
+size 152140
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..11ff4c136c33cbe7eb2579e50bbabe0c5532a967
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman with glasses is seen sitting inside a rustic wooden hut. She is holding a knife with a black handle and a white blade, and she appears to be sharpening the knife. The hut is made of wooden logs and branches, giving it a natural and earthy feel. The woman is dressed in casual clothing, and she seems to be focused on her task. The overall atmosphere of the video is calm and peaceful, with the woman and her surroundings creating a sense of tranquility."
+    ],
+    "video_ids": [
+        "5MUqsEhjar0_21_1236to1361"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with glasses, a knife with a black handle and a white blade.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing glasses, holding a knife with a black handle and a white blade. These elements are clearly visible and match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing glasses, holding a knife with a black handle and a white blade. The elements match consistent visible and match the description provided. the questionObject(s)' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..116fd26c4347802ff1e139c98cf506821e45354b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4281d997df559723285266d15bd6dc521eb46934a726414aaace1eb1e9d247b6
+size 163710
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..dd6cc5a13b736651b161cdf51f067abe5c925221
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a culinary tutorial, showcasing the preparation of a tomato-based soup. The style is simple and straightforward, focusing on the food and the process of cooking. The first frame shows a close-up of a bowl of soup, with a dollop of butter in the center. The second frame shows the same bowl of soup, but with the butter melted and swirled into the soup. The third frame shows the same bowl of soup, but with the addition of a garnish, possibly a sprig of basil or a drizzle of olive oil. The background is minimal, with a wooden table and a few tomatoes, emphasizing the freshness of the ingredients. The video is likely to be accompanied by a voiceover or text explaining the recipe and the cooking process."
+    ],
+    "video_ids": [
+        "x9xd9T8jYQs_10_0to155"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bowl of soup, dollop of butter, garnish (sprig of basil or drizzle of olive oil), tomatoes? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bowl of red soup with a visible dollop of butter melting on top, and a white swirl that appears to be a garnish (possibly cream or yogurt). Three whole tomatoes are visible in the background, fulfilling the requirement for tomatoes. The core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bowl of soup soup with a dol dollop of butter on on top. which a garn bowl that could to be either garnish,either olive or a). The tomatoes tomatoes and visible in the background, and the ' for tomatoes. The garn elements of in present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6e8df5d1d4dbca9bced69e09ad667c89678a7701
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24c795fbfb2b3e0ada18e92ee0bdc3370966983e1e721d5f6a057f9680744a9a
+size 89485
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..9d3e06cd6ddf71c051549607f719e42021d60d98
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man named Brett Larson, who is the host of a news segment for Buzz60. He is dressed in a suit and tie, and is standing in front of a newsroom with multiple monitors displaying various news stories. The monitors are turned on and displaying different images and text. The newsroom appears to be a professional and well-equipped environment, with several other people visible in the background, likely other news anchors or staff members. The style of the video is a typical news segment, with the host delivering the news in a professional and polished manner."
+    ],
+    "video_ids": [
+        "657OZnQ35qo_1_0to155"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Brett Larson (host), multiple monitors, other news anchors or staff members? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing Brett Larson (host) as the central figure, multiple monitors in the background, and other staff members working at desks, which aligns with the description. The presence of additional elements like lower third graphics does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showing showing Brett Larson,host) in the central figure. multiple monitors in the background, and no news members or in the. which ares with the description of The presence of the elements such the-third graphics and not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3327899d2d81f5a947e0745a6de283c8e3c6a57e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3808faa8e7a6d1e0b09508cd8d8a91d75f60f523a28a63d2d262008dabfec73
+size 129772
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..3337945558525876e7e00fe75d12f6ae64c7e0b9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse of a red and black motorcycle being assembled in a workshop. The motorcycle is adorned with various sponsor stickers, including \"MUNRO\" and \"SPIRIT OF INDIAN SCOUT\". The motorcycle is shown in three stages of assembly: first, the frame is visible with the wheels and engine in place; second, the body panels are attached; and finally, the motorcycle is fully assembled and ready for use. The workshop setting is filled with tools and equipment, and the motorcycle is the central focus of the video. The style of the video is a straightforward time-lapse, capturing the process of assembling the motorcycle in a clear and concise manner."
+    ],
+    "video_ids": [
+        "3xEJAmEd2Bg_21_0to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red and black motorcycle with sponsor stickers, various tools and equipment? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red and black motorcycle with sponsor stickers, including 'Spirit of Munro', 'Indian Scout', and various brand logos. The motorcycle is stationary on a stand, and while no tools or equipment are clearly visible, the core object (the motorcycle with sponsor stickers) is accurately represented. The presence of a small wheel stand is acceptable as it's part of the motorcycle's display setup.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red and black motorcycle with sponsor stickers, which logosSU' thester' whichM Motorcycle', and ' other logos. The motorcycle is positioned and a stand, and the there specific or equipment are visible visible in the setting description descriptionthe motorcycle) its stickers) is present depicted. The presence of a hand hand in and acceptable as it does a of the motorcycle setup setup setup.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..34ee2a896b3c0d576c4a496be24ac40518e347c0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c9d8a40ec0dea1a9c56b5cf64742bfdafa0517cbcf667ae241f3ec52fd0349e
+size 142839
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..29a5d6dc9278bac9692db7f8ab130f39aef7129a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment from a basketball game featuring the Golden State Warriors. The style of the video is a dynamic, action-packed sports sequence. The first frame shows the players on the court, with one player in the foreground looking towards the camera, while another player in the background is looking away. The second frame captures the same players, but with a different player in the foreground, who is also looking towards the camera. The third frame shows the players again, but with a different player in the foreground, who is looking away from the camera. The players are dressed in their team uniforms, which are blue and yellow. The court is filled with other players, and the atmosphere is intense and competitive. The video captures the energy and excitement of the game, as well as the camaraderie and teamwork among the players."
+    ],
+    "video_ids": [
+        "uJ2i1FZ-zvE_8_0to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Players on the basketball court? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three basketball players on what appears to be a court, wearing team uniforms and gear consistent with a basketball game or practice. Their attire, expressions, and positioning align with the description of 'Players on the basketball court'.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows basketball individuals players wearing a appears to be a basketball, wearing Golden uniforms. standing typical with a basketball game. practice. The attire and including, and the suggest with the ' of playersPlayers on the basketball court'. The}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..de6c658334417fa42dc3b2d81f021e4094a7ec6c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1120-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f8278ae70ba16d54b789f0ce4410cade22868fd2da49347c2ef663bc793534a
+size 331892
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..182607a3086ba4cf5af47a1f096fcc6ce13065da
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two wolves are seen interacting with a small doll in a nurse's outfit. The doll is placed on a white box with a red cross, suggesting a medical theme. The wolves are standing on a grassy area, and in the background, there is an old, rusted ambulance. The scene appears to be set in a forested area, with trees visible in the background. The overall style of the video is cartoonish, with vibrant colors and exaggerated features. The wolves and the doll are the main characters in this scene, and their actions suggest a playful or curious interaction."
+    ],
+    "video_ids": [
+        "K3FrCBNwRT8_214_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two wolves and a small doll in a nurse's outfit are the main objects in the scene. There is also a white box with a red cross and grassy ground.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts two wolves and a small doll dressed as a nurse as the main subjects. A white box with a red cross is clearly visible on the ground, and the scene is set on a grassy ground with trees in the background. All core elements from the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful two wolves and a small doll dressed as a nurse. the main objects. The white box with a red cross is also visible, the ground, and the background is set on a grassy area. an and the background. The elements elements of the description are present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a6955e48249216d5fab9451994e7cf4daf3c32fd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:effe2fc48f1ba3553c8cbf0534a003c366a610e2f19a19b228ade0e1059cc8eb
+size 150920
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..59aa30f73c45e94d4f987835113ff179dbd7ce40
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man with a beard and a blue shirt taking a selfie. He is wearing a cap and has a backpack on. The background features a stone building and a tree. There are other people walking in the background. The man is standing on a sidewalk. The style of the video is casual and candid, capturing a moment in the man's day."
+    ],
+    "video_ids": [
+        "72Mpxf13zjM_51_489to611"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and a blue shirt, wearing a cap and a backpack.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man with a beard and a blue shirt, wearing a cap and a backpack, which matches the description. Additional elements like other people and buildings in the background do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man with a beard and a blue shirt, wearing a cap and a backpack. which matches the description provided The elements like the people in a in the background do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..68b88f29d020f368d465342cabaa425651a0f6b1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1bd78dc7a44e0c577a4605ada2ac0c2a16d8011a11abe1b116ef9678f5d82a06
+size 278888
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..c568d398b183a028fcfd0eac8eef6aec2af4d2e8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a hard drive with a silver chain and a gold padlock securing it. The chain is wrapped around the hard drive, and the padlock is locked, indicating that the hard drive is protected and secure. The style of the video is straightforward and clear, focusing on the hard drive and the security measures in place. The video does not contain any additional objects or actions, and the focus is solely on the hard drive and its security features. The video is likely intended to convey the importance of data security and the measures taken to protect sensitive information."
+    ],
+    "video_ids": [
+        "v0QkafslnrM_22_0to164"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Hard drive, silver chain, gold padlock? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a hard drive with a silver chain wrapped around it and a gold padlock securing the chain, fulfilling the described objects. The text overlay clarifies that this is a metaphor, but the core objects are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a hard drive with a silver chain and around it and a gold padlock attached the chain. which the ' objects. There presence ' mentioningifies the the is a 'ical but the visual visual in present represented.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..24f1da1ec1efd0be4657c73873a9fe98375f0b65
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:97497bf8225bf3a1a9e0a45a2391a4888ac567de4602d7e4a288b14e08436d6c
+size 224047
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..1de0bb4d027b0f47991500c618bf1b33e00784ea
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red sports car with the word \"ADVANTAGE\" written on its side, driving on a track. The car has a sleek design with gold rims and a black roof. The track appears to be a race track with other cars parked in the background. The car is moving at a high speed, as indicated by the blurred background. The style of the video is dynamic and action-oriented, capturing the thrill of racing."
+    ],
+    "video_ids": [
+        "QUMGSXL5QhE_12_0to134"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red sports car with 'ADVANTAGE' written on its side, gold rims, and a black roof.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red sports car with 'ADVANTAGE' written on its side, gold rims, and a black roof, matching the description. Additional cars and background elements are present but do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts a red sports car with 'ADVANTAGE' written on its side, which rims, and a black roof. which the description provided The elements and a elements are present, do not contradict the core description of}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..30849526d66decf6cc1ef2e968842805d3e4b0f9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cecc351a171bd6cb2ca322b884147366ca573fa855d2e4723ea99f8641ea2a96
+size 175457
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..0a36cf619d4aa09291499af0192c8582d493428d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a man sitting at a cluttered desk in an office setting. He is wearing glasses and a plaid shirt, and he appears to be in deep thought or possibly experiencing frustration. The desk is filled with various items, including papers, binders, a phone, and a cup. Behind him, there is a wall with a patterned wallpaper, and a window with blinds. The overall style of the video is realistic and it seems to be a scene from a daily office life."
+    ],
+    "video_ids": [
+        "RF0E0rkZXAk_7_26to148"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, papers, binders, a phone, a cup? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man sitting at a desk surrounded by papers, binders, a phone, and a cup (a pen holder with pens and pencils). All the specified objects are present and visible in the scene, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man sitting at a desk with by papers, binders, and phone, and a cup.which mug holder). a and a). The the objects objects are present and the in the scene, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1243cc59bec6de051cbfc10a3bab8f7dcd3b7a0c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:02f149d3719bca59c2bfc12b608851ffcd1225118f8c9d72d42add150f88d346
+size 173287
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..ad001d5ae6686adff22f95a8f223c4893fe24e58
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a blue Jaguar car driving on a gravel road. The car is sleek and modern, with a shiny exterior that reflects the sunlight. The road is surrounded by a natural landscape, with mountains in the distance and a clear blue sky overhead. The car is moving at a moderate speed, and the driver is focused on the road ahead. The overall style of the video is dynamic and adventurous, capturing the thrill of driving in a beautiful and challenging environment."
+    ],
+    "video_ids": [
+        "XJOgDZUklfs_9_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue Jaguar car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a blue Jaguar car parked on a gravel surface with a scenic mountain and cloud background. The car's design, color, and branding are consistent with a Jaguar model, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a blue Jaguar car, on a rocky road with a mountain mountainous forest-filled. The car is design and color, and the are consistent with the Jaguar model, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..339033dc20456f73746eff80a4987fae07ed8469
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a2fdf46ddac6ac55a348c776254267f851aafab6c725f3be1da931f18311798
+size 339673
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..ae49206e32bf1f93b2a2a0150a7d12da93cdc84e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man with a beard and short hair, wearing a black leather jacket and a dark shirt. He is standing in a parking lot with a gray Audi car behind him. The man appears to be speaking or reacting to something, as he is looking to the side with a slight smile on his face. The parking lot is paved with gray bricks, and there are trees and a fence in the background. The lighting suggests it is daytime. The style of the video is casual and candid, capturing a moment in the man's day."
+    ],
+    "video_ids": [
+        "O6VCabmobqY_40_22to173"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and short hair, wearing a black leather jacket and a dark shirt, and a gray Audi car.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and short hair, wearing a black leather jacket and a dark shirt, standing next to a gray Audi car. The description matches the visual content without contradiction, even though there are additional background elements like trees and snow.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and short hair, wearing a black leather jacket and a dark shirt. standing in to a gray Audi car. The description and the core elements of any or and though the are additional elements elements like trees and a that}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a73aabb3065dac0dadd5b8ddaf61a8c6dd4b6915
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d9675d8dc7f8fcbc7894b4a1e5ae85b6f8d78ca48669d9445df66c9c222e1770
+size 145774
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..0be7c48c419608e2ea87ace8ec94c5d4079a2fe7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment from CNN, focusing on the first 100 days of the Trump administration. The style of the video is a mix of live footage and news commentary. The live footage shows President Trump signing a document at a desk in the Oval Office, surrounded by his advisors. The advisors are dressed in suits and ties, and they are standing behind the President, watching him sign the document. The news commentary is overlaid on the live footage, providing context and analysis of the event. The commentary discusses concerns over the Trump administration's use of \"alternative facts\". The overall tone of the video is serious and informative."
+    ],
+    "video_ids": [
+        "NRjIZwg879s_15_0to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: President Trump, his advisors? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows President Trump seated at a desk, surrounded by several men in suits who appear to be his advisors. The setting, attire, and actions (e.g., signing a document) are consistent with the description of President Trump and his advisors. The presence of additional elements like the CNN news banner and background flags does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts a Trump seated at a desk, signing by individuals individuals and suits, appear to be his advisors. The setting and attire, and the alignsuch.g., signing a document) align consistent with a description of a Trump and his advisors. The presence of additional elements like the American logo ticker and the figures does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c2c997442304e6f054108db1000628d587ef10da
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1121-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04ca7172d4a906dc3ea6d5b8a09d836b6f6e456a3d61c8e41d9b79d6ee47ce12
+size 190845
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..c8d83be574a70516f49102b9bbe9928aa23e7089
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man sitting on a stone bench in a park-like setting. He is wearing a brown jacket with colorful patches on the sleeves, blue jeans with a ripped knee, and glasses. The man has a beard and is looking off to the side with a thoughtful expression. The background features a stone archway and lush greenery, including trees and bushes. The lighting suggests it is daytime. The style of the video is casual and candid, capturing a moment of quiet reflection in a serene outdoor setting."
+    ],
+    "video_ids": [
+        "ZQEMiRlTLho_51_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man sitting on a stone bench, wearing a brown jacket with colorful patches on the sleeves, blue jeans with a ripped knee, and glasses. The man has a beard.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting on a stone bench, wearing a brown jacket with colorful patches on the sleeves, blue jeans with a ripped knee, and glasses. He also has a beard, matching all the specified details in the description. The background elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting on a stone bench, wearing a brown jacket with colorful patches on the sleeves, blue jeans with a ripped knee, and glasses. The also has a beard, which the the details details in the description. The background includes, not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..29347dc84c5e7ef6416fbd0678c955a0c1b6c1aa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8dd00a7dfab69355534ad855cb9524a492a8d0b4d7a70049ad562bae3098ba7c
+size 152042
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..d3ea156a60a1a242998884f25f421a2b39ecc40a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a scene from a Korean TV show or movie. It features a woman standing in a rustic outdoor setting, possibly a garden or a farm. She is wearing a floral apron over a striped shirt, suggesting she might be involved in some sort of outdoor activity or work. The woman is looking off to the side, her expression is one of concern or worry. The background is filled with greenery and wooden structures, indicating a rural or semi-rural environment. The overall style of the video is naturalistic, with a focus on the woman and her immediate surroundings. The lighting is soft and diffused, suggesting an overcast day or a shaded area. The colors are muted, with a predominance of greens and browns, which adds to the rustic feel of the scene."
+    ],
+    "video_ids": [
+        "6ZO9k7tKeMU_27_0to196"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a floral apron over a striped shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The woman in the video is clearly wearing a floral apron over a striped shirt, matching the described attire. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video in the video is wearing wearing a floral apron over a striped shirt, which the description '. The background and additional elements in not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..54d114c2fd1c559c7a3619a94d87624bc955a024
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2da37a47ec2257912e120158a8b56fa28656a616751b3ac08cccd6a947e174b0
+size 253930
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..b3f112d83c0172418c07455c5e56729308a2a32b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a blue SUV parked on a gray surface. The car is positioned at a slight angle to the camera, allowing a clear view of its side profile. The vehicle features a sleek design with a prominent grille and sharp lines. The wheels are silver, and the car has a glossy finish. The background is neutral and does not distract from the car. The style of the video is a straightforward product showcase, focusing on the car's design and color. The lighting is even, ensuring that the car's details are clearly visible. The video does not contain any text or additional elements."
+    ],
+    "video_ids": [
+        "kR4WiWRWDLI_13_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Blue SUV, silver wheels, sleek design with a prominent grille and sharp lines, glossy finish? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a blue SUV with silver wheels, a sleek design featuring a prominent grille and sharp lines, and a glossy finish that reflects light. All elements described in the condition are clearly visible and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a blue SUV with silver wheels, which sleek design featuring a prominent grille, sharp lines, and a glossy finish. matches light, These these of in the condition are present visible and match represented in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7dae23f694bd201da57e4d7492674e5a09e9a1e3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc53b8aadb2649398934cdcc0e0949a657952eb7a2186bd9a67215b48b9aa72a
+size 133964
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..ed999f55c7d18ad27d8806d925366d2f0b1694fd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen interacting with a horse in a serene outdoor setting. The woman, dressed in a black jacket, stands next to the horse, smiling as she gazes at it. The horse, with its dark brown coat, stands calmly, its head turned towards the woman. The background features a blurred landscape, suggesting a peaceful, natural environment. The overall style of the video is a close-up shot, focusing on the interaction between the woman and the horse, with the background serving as a subtle backdrop to the main subjects."
+    ],
+    "video_ids": [
+        "Gj642Uy1qnI_5_105to227"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman and a horse? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman wearing a black jacket and a horse with a bridle. The woman is interacting with the horse, which aligns with the 'Object(s)' condition of a woman and a horse. Additional elements like a stone wall and lighting equipment in the background do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a woman and a black jacket and a horse with a whiteidle. The woman is standing with the horse, which iss with the 'Object(s)' condition of the woman and a horse. The elements such the grass wall in a do in the background do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f6a410f82a083f648e2485fbae78bf6cc8841c21
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6fae95c248031039ad58fbe3e18d01c47205d7ac1abeb411e75e8ac96bfc260
+size 114185
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..707fa175f203ce8276be2d832f547d9f0bb7e791
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a dynamic and colorful depiction of a motorcycle race. The scene is set on a winding track with a lush green hillside in the background. The track is lined with blue and white banners, adding to the vibrant atmosphere.  In the first frame, a red Ducati motorcycle is seen from the rider's perspective, giving a sense of speed and motion. The motorcycle's sleek design and vibrant red color stand out against the green backdrop.  In the second frame, the same red Ducati is seen from a side angle, showcasing its aerodynamic shape and the rider's protective gear. The motorcycle's position on the track suggests a high-speed turn.  In the third frame, the red Ducati is seen from a rear angle, emphasizing its powerful engine and the rider's skill in maneuvering the motorcycle. The motorcycle's position on the track indicates that it is in the midst of a thrilling race.  Throughout the video, the motorcycles' sleek designs and the riders' protective gear are clearly visible, adding to the overall excitement and intensity of the race. The lush green hillside and the winding track provide a stunning backdrop for this high-speed event."
+    ],
+    "video_ids": [
+        "xGepyBQy86g_49_0to135"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red Ducati motorcycle, green hillside, blue and white banners? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a red Ducati motorcycle prominently in the foreground. A green hillside is visible in the background behind the fence. Blue and white banners are also present in the background, though slightly blurred, confirming their presence. Additional motorcycles are visible but do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a red Ducati motorcycle in in the foreground, The green hillside is visible in the background, the motorcycles, Blue and white banners are also present along the background, consistent they less, they the presence. The elements and visible, do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d6bfb2095af9fb46697ca96df6be2c6506dde475
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c12bd9978a061c69f9507c307ccbc84fd4c550976721562702c8151863f20140
+size 243020
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..cc846d0edbdc2054ff52749843e4028021122769
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a blue BMW car parked on a snowy road. The car is positioned in the center of the frame, with the rear facing the camera. The car's design is sleek and modern, with a prominent rear bumper and a distinctive taillight design. The car's license plate reads \"YX18 KVY\". The road on which the car is parked is covered in a thin layer of snow, and there are mountains visible in the background. The sky is clear and blue, suggesting a cold winter day. The overall style of the video is realistic and naturalistic, with a focus on the car and its surroundings."
+    ],
+    "video_ids": [
+        "LejmPFmRbIA_19_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue BMW car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a blue BMW car, specifically a BMW M5, from a rear three-quarter angle. The car's distinctive design, including the taillights, exhaust, and BMW logo, matches the description. The surrounding environment (mountains, road) does not contradict the core object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a blue BMW car, which a BMW M2, which the rear three-quarter view. The car's distinctive features, including the taillights and rear, and overall logo, is the description of The presence environment,snowains, snow, and not contradict the core description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..671007f902d1c8e7ca7a9a4b662985918aa5f86e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc2a3885349070c253bc2a96dfababec15107426b3d2a2ca0017aa8b7405b7d7
+size 158496
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..df44cdb831f0f9c66a9b2adaca300352bb0432df
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene sunset scene at a beach. The sun is positioned low on the horizon, casting a warm golden glow across the sky and reflecting off the calm water. The sky is painted with hues of orange, yellow, and pink, transitioning into a darker blue towards the top. Clouds are scattered across the sky, some illuminated by the setting sun, adding depth to the scene.  In the foreground, tree branches with green leaves hang down, framing the view and adding a natural element to the composition. On the left side of the frame, a small dog is seen walking along the shoreline, moving from left to right. The dog's presence adds a sense of life and movement to the otherwise tranquil setting.  The water is smooth and still, mirroring the colors of the sky and creating a beautiful reflection of the sunset. The overall atmosphere is peaceful and calming, with the natural beauty of the sunset and the gentle movement of the dog providing a soothing visual experience. There are no significant changes"
+    ],
+    "video_ids": [
+        "956065bc398fcbf7e0f6168f2e48fbd95106845542421dff79a242cf70c12294"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Tree branches with green leaves, a small dog? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows tree branches with green leaves framing the top of the scene, and a small dog is visible on the beach near the water's edge, matching the described objects. The presence of the sunset and ocean does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts tree branches with green leaves on the left left the scene, and there small dog is visible on the beach. the water. edge. which the description objects. The presence of the sunset and the does not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cb028c78ef5fa042b2eb840bc1508652d145249d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1f511a1aa1128d624c5ac0b17c274fdf8f93a91d2a6a61bdc4940e9df236622f
+size 163781
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..de40dd91b1b54a906a5ad84b3e53000a72d400a0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is a player wearing a black and teal uniform with the number 27, who is in possession of the football. He is running towards the right side of the frame, with his body leaning forward in a sprint. His helmet is black with a teal visor, and he is wearing white gloves.  In the background, there are other players, some of whom are wearing white uniforms with red accents. They are in various positions on the field, some closer to the camera and others further away. The field itself is a vibrant green, and the sky is clear, suggesting it's a sunny day.  The style of the video is a standard sports action shot, capturing the intensity and speed of the game. The camera angle is slightly elevated, providing a clear view of the player's movement and the surrounding action. The focus is sharp on the player with the ball, while the background is slightly blurred, emphasizing the speed and movement of the scene."
+    ],
+    "video_ids": [
+        "gwDydZmCcyw_43_0to131"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a black and teal uniform with number 27, other players in white uniforms with red accents, and the football.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a black and teal uniform with the number 27, holding a football. Another player in a white uniform with red accents is visible nearby, consistent with the description. The core elements described are accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a black and teal uniform with the number 27, which a football. There player in a white uniform with red accents is visible in, and with the description. The setting elements of are present represented in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ee15b6b8692a8538b30ada31e24c9e74142f17c5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1122-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce947ca6de8ef568c9034c9e95e3376fec7abccddae195cd51387562788e7502
+size 238346
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..8eac1e0938b4d45191a9f0daad1745eb61c27b99
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows an orange and white cat eating from a gray bowl on a wooden floor. The cat is wearing a collar with a purple tag. The cat is sitting on the floor and is focused on the food in the bowl. The bowl is filled with cat food. The cat's fur is a mix of orange and white. The wooden floor provides a warm and natural backdrop for the scene. The cat's collar adds a touch of color to the otherwise neutral tones of the scene. The cat's focused attention on the food suggests that it is enjoying its meal. The overall style of the video is simple and straightforward, capturing a common everyday moment with a sense of warmth and comfort."
+    ],
+    "video_ids": [
+        "8A47l2_KQpk_4_0to187"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An orange and white cat, a gray bowl, and a purple collar with a tag.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an orange and white cat eating from a gray bowl, and the cat is wearing a purple collar with a tag, which matches the described objects. There are no conflicting elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an orange and white cat eating from a gray bowl. which the cat is wearing a purple collar with a tag. which matches the description elements. There are no additional elements in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..825f23ef1a9b6f4f9cf19d9759a782122f2f2d1e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b63667276e1c8b109d7dcfb31375daf5fe94848d0982c3e613a6b65be335fcf
+size 168346
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..5007e818867473737c9fd36c3fa19b95e3922e26
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and short hair, wearing a dark shirt. He is seated in a studio setting with a blue background. The man appears to be engaged in a conversation or interview, as suggested by his attentive expression and the presence of a microphone. The text overlay on the video reads \"Steve Wilkos It's your baby, are you back on drugs?\" This suggests that the video is likely a segment from a talk show or a similar television program, where the host, Steve Wilkos, is addressing a serious topic related to drug use and childcare. The style of the video is typical of a television talk show, with a focus on the host and the guest, and the use of text overlays to convey important information."
+    ],
+    "video_ids": [
+        "BGlRpEbaf5A_6_108to290"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and short hair wearing a dark shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and short hair wearing a dark shirt, which matches the description. The background and text overlay do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and short hair wearing a dark shirt, which matches the description provided The background and the overlay are not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ab2b69a7529baaa620db415f61f66a886889a2b3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9b4ed330e9a6e5f263d5cb468d434ae7012e17a979bf17f5e640d1359528b6e4
+size 135366
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..315af84607d2649425fd307a79a1362788e74968
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a drone shot of a river flowing through a valley with steep, rocky cliffs on either side. The river is brown and appears to be shallow, with exposed rocks and sandbars. A dirt road runs parallel to the river, with a few people walking along it. The vegetation on the cliffs is sparse, with patches of grass and shrubs. The sky is overcast, and the lighting is soft, suggesting either early morning or late afternoon. The overall style of the video is naturalistic, capturing the beauty of the landscape without any human intervention or artificial elements."
+    ],
+    "video_ids": [
+        "3WWnYwz8DMw_18_185to336"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: River, rocky cliffs, dirt road, people walking, vegetation (grass and shrubs)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a river with muddy water, rocky cliffs on both sides, a dirt road running alongside the river, people walking on the path, and vegetation including grass and shrubs. All core elements described are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a river flowing a water, rocky cliffs on either sides, and dirt road running parallel the river, and walking along the road, and vegetation such grass and shrubs. The these elements of in present in accurately with the video content.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a9cdb3c61c79716867178ad157f7951835e3a269
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:225fa073181dc53bde4ce57b338f933c418a2f595a634cae4df7ccc2fb9b693c
+size 157101
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..e56043c7536bb449844370dfb67f5fddb856b8c7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man standing at a podium in a room with American flags on either side. He is wearing a suit and tie and appears to be speaking. The room has a fireplace and a clock on the wall. The man is the main focus of the video, and the setting suggests a formal or official event. The style of the video is straightforward and documentary-like, capturing the man's speech without any additional embellishments or distractions."
+    ],
+    "video_ids": [
+        "1wnZRs12rn8_2_972to1099"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie standing at a podium, giving a speech.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit and tie standing at a podium with a microphone, which aligns with the core description of giving a speech. The setting includes American flags and formal decor, which are consistent with a public address, and do not contradict the main condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit and tie standing at a podium, an microphone, which iss with the description description of giving a speech. The presence includes an flags and a decor, which are consistent with a formal or or possibly the not contradict the description action.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bad3f02104a0bef789819979188ef7222be05ccb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:763ca871978357c76e54c010ba8a2ece9d7b6bc364a8ce8756a3c259c5160ee2
+size 80874
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..014b4c53ff4be0cc6f281bfff5bf4c428fcadd21
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a meal in progress, with a focus on the table setting and the food being consumed. The table is set with a variety of dishes, including bowls, plates, and a glass of water. The food appears to be a mix of Indian cuisine, with dishes such as naan bread, chutney, and curry. The table setting includes silverware, including forks, knives, and spoons, as well as salt and pepper shakers. The style of the video is a simple, straightforward documentation of a meal, with no additional embellishments or artistic effects. The focus is on the food and the table setting, with no additional context or narrative provided. The video is likely intended to showcase the meal and the dining experience, rather than to tell a story or convey a specific message."
+    ],
+    "video_ids": [
+        "It9QNXobD4o_116_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bowls, plates, naan bread, chutney, curry, glass of water, forks, knives, spoons, salt shaker, pepper shaker? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows all the listed objects: bowls (with remnants of food), plates (one with a knife, another with remnants), naan bread (in a bowl), chutney (in a small container), curry (in the rectangular plate with residue), glass of water, forks, knives, spoons, and salt and pepper shakers. The scene is consistent with a meal setting, and all requested items are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful various the objects objects: bowls,with food of food), plates (with with a small and one with a of naan bread,being the bowl), chutney (in a small bowl), curry (in a bowls dish), a), glass of water ( forks, knives, spoons, salt salt and pepper shakers. The presence depicts set with a meal setting, and the elements objects are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..082899006bf11c8efb064a2a5ac31e6a3ff8e464
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:65922a8d458a5da739482ac542365c1eae7d7111e8ba589b23b44ae418368326
+size 129292
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..52c63e5d1755240528079b165029bd100df985d8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a black poodle dog standing on a rock near a body of water, possibly a lake or a river. The dog is on a leash and appears to be looking at the camera. The background features several boats docked at a marina, with the water reflecting the boats and the sky. The style of the video is a simple, straightforward, and unedited capture of a moment, likely taken by a pet owner or a passerby. The focus is on the dog, with the background serving as a natural and serene setting."
+    ],
+    "video_ids": [
+        "sYejT69d9xg_2_0to169"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black poodle dog, a rock, a body of water, several boats, and the sky.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a black poodle dog standing on a rock, with a body of water and several boats in the background. The sky is not clearly visible, but the presence of the water and boats implies an outdoor setting with sky, which is acceptable as per the condition. The core elements described are all present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a black poodle dog standing on a rock by with a body of water in several boats in the background. The sky is not directly visible in but the lighting of the boats and boats suggests it outdoor setting that a in which is acceptable as long the description. The dog elements of are present present in match represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..15cc8e10c629a0320b7db320269c71501698ca7b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f83f7b7dd744f558aa0f788f9221a88cb7b533a3fa785822017716801f12391
+size 192176
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..317a551ffe0c9029d4d0e922d9c2b06537cc45ec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment from a football game, featuring a player from the Cleveland Browns team. The player, wearing a black jersey with the number 10, is seen sitting on the ground, possibly after a play. The image is slightly blurred, indicating motion or a fast-paced game. The player's helmet is off, suggesting that the play has ended or is paused. The background is filled with other players, some in motion, indicating an ongoing game. The style of the video is dynamic and action-packed, capturing the intensity of the sport."
+    ],
+    "video_ids": [
+        "bS6QNeqL9Po_0_0to122"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player from the Cleveland Browns team, wearing a black jersey with the number 10, and other players in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a Cleveland Browns uniform with the number 10 on his black jersey, sitting on the field. Other players are visible in the background, some in Browns uniforms and others in opposing team uniforms, which aligns with the description. The scene is consistent with an American football game setting.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a player wearing a black Browns uniform with the number 10, the jersey jersey, which on the field. The players in visible in the background, which wearing similar uniforms and others in different team uniforms, which aligns with the description. The setting is consistent with a NFL football game setting.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..159ff31fcff1544fa636e0721fda214ecaeec1da
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9b120f01cbb8f048ca2b3234f7a6dc45ccf0dd1a30915f941c8809743608d4ee
+size 251371
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..a80df79127a63fc7f14e2037b85c76e5100db87c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up of a tiger's face in three frames. The tiger's fur is a mix of black, white, and brown stripes, and its eyes are a striking amber color. In the first frame, the tiger's eyes are wide open, looking directly at the camera. In the second frame, the tiger's eyes are slightly narrowed, giving it a more intense look. In the third frame, the tiger's eyes are fully closed, and its mouth is slightly open, as if it's about to roar. The style of the video is realistic and detailed, capturing the majestic beauty of the tiger in its natural habitat."
+    ],
+    "video_ids": [
+        "FZ5xUmzXSVc_40_0to130"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Tiger's face? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features the face of a tiger in close-up, which matches the 'Object(s): Tiger's face' condition. Although another tiger's face appears partially in the background, it does not contradict the core description and is acceptable as an additional element.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a face of a tiger, a-up, capturing align the 'Object(s)' Tiger's face' condition. The the tiger is face is in in the background, the does not detr the main description as is acceptable as additional additional element.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..169b74b5ab55681c1965be6c5ef6e4f1a4f83722
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1123-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1613e8308de5614aa43aa787d551832efd1883409a6c8d841c302c256e8056e8
+size 407343
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..ad6cdb3b14506a6d8075a94fc6afcaca1cb45b8c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with long brown hair styled in a bun, wearing a red off-the-shoulder top. She is seated in a studio setting with a brick wall in the background and a potted plant on a shelf. The woman appears to be engaged in a conversation, as suggested by her attentive gaze and slight smile. The overall style of the video is a casual, yet professional interview or discussion, with the woman as the main subject. The lighting is soft and even, highlighting the woman's features without creating harsh shadows. The studio setting is minimalistic, with the focus on the woman and her interaction."
+    ],
+    "video_ids": [
+        "VLC-aJz7hkI_6_0to161"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with long brown hair styled in a bun, wearing a red off-the-shoulder top, sitting attentively and smiling slightly during a conversation.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The woman in the video has long brown hair styled in a bun, is wearing a red off-the-shoulder top, and is seated while engaging in conversation with a slight smile. These elements match the description provided, and the background elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video in the video has long brown hair styled in a bun, wearing wearing a red off-the-shoulder top, and appears sitting in smiling in a. a slight smile. The elements match the description provided, indicating there presence,, not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f3ffa0a69103c5b5c422e8e47df6dd4cb0a1b1ff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c0f0def88e1f56ff171e62e6678a68ed32fd6bc787daa50dfa6bcfb1adcc59ff
+size 138759
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..0e19bf13e37df40b5aa6d01c54c657fefe65d70b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man sitting at a desk in a room with a blue background. He is wearing a white t-shirt and has a beard. In front of him is a laptop. Behind him, there is a shelf with various objects, including a skull, a globe, and a model of a spaceship. The man appears to be speaking or presenting, as he is gesturing with his hands. The style of the video is casual and informal, with a focus on the man and his surroundings."
+    ],
+    "video_ids": [
+        "05LRfAD5tqk_67_21to186"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man, laptop, shelf, skull, globe, model of a spaceship? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting at a desk with a laptop in front of him. Behind him is a shelf with various items, including a skull in a glass dome, a globe, and a model of a spaceship (which appears to be a stylized or abstract representation of a spacecraft). All the specified objects are present, and there are no contradictions with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting at a desk with a laptop, front of him. Behind him, a shelf with a objects, including a skull, a helmet case, a globe, and a model of a spaceship.which appears to be a smallized or artistic representation rather a spaceship). The the objects objects are present and and the are no contradictions with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d2ed4bc9d5df38094f6a5382fdc9e27e5835b0ed
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0da48e5b25a345e2975d89ffeb0ef36d2f1fb9ef3fa75b341e1f595a71a3a906
+size 141112
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..5bbaba3727f6f19d19feb112e0347cc435bdb0f3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a soccer match. A player, dressed in a vibrant red jersey and shorts, is in mid-air, executing a powerful kick with his right foot. The player's body is angled towards the right side of the frame, suggesting a strong, forward-moving force. The player's left arm is extended for balance, while his right arm is bent at the elbow, indicating the follow-through of the kick. The background is a blur of spectators, their faces a mix of anticipation and excitement, adding to the intensity of the moment. The focus is solely on the player and his action, creating a sense of motion and energy. The style of the video is realistic, capturing the details of the player's attire and the spectators' expressions with clarity. The overall composition of the video suggests a high-stakes moment in the game, with the player's kick being the central action."
+    ],
+    "video_ids": [
+        "bv5R0ED4FYI_15_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a vibrant red jersey and shorts, with spectators in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a vibrant red jersey and shorts, consistent with the description. The background clearly features spectators, fulfilling the 'Object(s)' condition. The player's attire and the setting match the core description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a player player wearing a vibrant red jersey and shorts, which with the description. The player includes features spectators, which the 'Object(s)' condition. The player's attire and the presence are the description description provided any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dc77372ee1f6955d349ad99e7e51db1d207aed1f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2113e2f84a38ff32be9c2e58c758962ede694eccbb3a927a1492510ed209e31a
+size 130478
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..a5c4225f23a1605501d9721e93a8c1d1124a62ce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a blue shirt with a mustache, engaged in a conversation with another person whose face is not visible. The man appears to be speaking, with his mouth slightly open and his eyes focused on the other person. The setting appears to be an indoor room with warm lighting, possibly a living room or a similar domestic space. The style of the video is a close-up shot, focusing on the man's face and upper body, which suggests a personal or intimate conversation. The image quality is clear, with good lighting and color saturation, indicating that the video is likely professionally shot or edited."
+    ],
+    "video_ids": [
+        "JlVYr3olXqE_606_0to186"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue shirt with a mustache, another person (face not visible).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a blue shirt with a mustache, and another person whose face is not visible, matching the description. The background and lighting are consistent with a scene from a film, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue shirt with a mustache, and there person is face is not visible. which the description provided The presence is the are consistent with the typical where a video or and there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0ae9d41df51e64c9cff2733edb687e348aa8e252
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c45f8e1a3c4a757eb77d8f143c284293ba677c3784c06165ac406472a6c9705
+size 141810
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..34f22aeb7b2fb7c6f68239e2db9f10bb76a7c416
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a fierce and menacing creature, possibly a demon or a dragon, in a fiery and hellish environment. The creature is predominantly red and black, with sharp, jagged teeth and a menacing expression. Its body is muscular and powerful, with a large, imposing head and a pair of wings that are spread wide. The creature's eyes glow with an intense, fiery light, and its skin appears to be scorched and charred, adding to its intimidating appearance. The background is a dark, fiery landscape, with flames and smoke billowing around the creature, creating a sense of chaos and destruction. The overall style of the video is dark and intense, with a focus on the creature's menacing presence and the fiery, hellish environment."
+    ],
+    "video_ids": [
+        "-NLRbowLfH0_47_64to271"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A fierce and menacing creature (possibly a demon or dragon) with sharp, jagged teeth, a large head, muscular body, and wide-spread wings.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a demonic creature with sharp, jagged teeth, a large head, and a muscular body. It also has wide-spread wings, matching the description. The creature's menacing appearance is emphasized by its glowing eyes and fiery red coloration, fitting the 'fierce and menacing' criteria.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a creature creature with sharp, jagged teeth, a large head, and a muscular body. The has has wide-spread wings, which the description of The fiery's appearance appearance, enhanced by the glowing eyes and the background and,, which the 'fierce and menacing' aspect.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4ee02570feb3001dc1af61bbe884e884632dbec3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2339d871f2db42214939622ccd078a6db26bf44d56eeeb16372a9fa0efa4543e
+size 280973
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..654a2d9d5284533cfd308806bd406c9ce6a76be5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a bald man with glasses, wearing a blue denim jacket and a pink tie, sitting in a living room with a window and a potted plant in the background. He is engaged in a conversation with another man who is not visible in the frame. The man in the denim jacket is identified as Ken Wingard, a DIY expert, and his social media handle is provided as @KenWingard. The video is likely a segment from a television show or a web series, as indicated by the logo of the Hallmark Channel in the bottom right corner. The style of the video is casual and informative, with a focus on the man's expertise in DIY projects."
+    ],
+    "video_ids": [
+        "lzGLe6NTt5A_2_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bald man with glasses wearing a blue denim jacket and a pink tie, and an unseen man engaged in a conversation.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man with glasses, wearing a blue denim jacket and a pink tie, who is engaged in conversation. Another man is partially visible on the left side of the frame, indicating an ongoing dialogue. The core description is accurately represented, with no significant contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man wearing glasses wearing wearing a blue denim jacket and a pink tie, seated appears engaged in a with The person is partially visible, the left side of the frame, suggesting that unseen conversation. The setting elements is largely represented in and no significant contradictions or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3ea985934496804cbfb806b189dc40a82683bd69
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b02989d9f6ed9331342423bd74f417447824d07e836dff1ddcf603a292cf4e2a
+size 214752
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..479c09632429f20be1b042cb9f178b454ea647b6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a collection of coffee mugs and containers on a table. The mugs are of various sizes and colors, including red, white, and gray. Some mugs have handles, while others do not. The containers are white and have lids. There are also spoons on the table. The style of the video is a simple, straightforward presentation of the items without any additional context or action. The focus is on the objects themselves, and there is no movement or interaction between them. The lighting is even, and the colors are natural. The video does not contain any text or narration."
+    ],
+    "video_ids": [
+        "vIZxL0R_s1o_41_119to257"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Coffee mugs (various sizes, colors: red, white, gray), containers (white with lids), spoons? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple coffee mugs in various sizes and colors (red, white, gray), white containers with lids, and spoons, all of which match the described 'Object(s)' condition. Additional elements like cocoa containers and coasters are present but do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a coffee mugs of various sizes and colors,red, white, gray), containers containers with lids, and aons. which of which match the description objectsObject(s)' condition. The elements like a mix and aers are present but do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d759faca5829497dd3758cabbdf3f601c835fe4f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e10b6162c70a58094382390dc2cb609555f16502fa7915eae3ae9157679e71e
+size 87549
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..467c450092768bc2fa51aa746cdb8ce7100f3251
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a baseball player in a New York Yankees uniform, standing on a baseball field. The player is wearing a cap and sunglasses, and appears to be in the middle of a game or practice. The field is well-maintained with clear lines and a well-defined diamond. The player is the main focus of the video, and there are no other significant objects or people in the frame. The style of the video is realistic and captures the intensity of the sport."
+    ],
+    "video_ids": [
+        "etDN6x64F28_3_104to278"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A baseball player in a New York Yankees uniform, wearing a cap and sunglasses.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a New York Yankees uniform, a cap with the Yankees logo, and sunglasses resting on the cap. These elements match the description provided in the 'Object(s)' condition. The background and other details do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a baseball wearing a New York Yankees uniform, including cap, the Yankees logo, and sunglasses. on his cap. The elements match the description provided, the questionObject(s)' condition. The man and setting details in not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5e2b088641cca8521c2645b0a9be5b629369e85c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1124-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb91b24eb489368cbc8ef8674786a68bb0821eb35c21e0b1cb2b313f50635f68
+size 166613
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..1aa6e45c4ba523a384fd56dfac81e231b3116c67
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a red sports car, showcasing its luxurious design and advanced technology. The car's interior is predominantly black, with red accents adding a touch of sportiness. The steering wheel, adorned with the brand's logo, is positioned in the center of the dashboard, which houses a digital display and various control buttons. The car's seats are upholstered in black leather, providing a comfortable and stylish seating arrangement. The car's interior is well-lit, with natural light streaming in through the windows, highlighting the car's sleek design and high-quality materials. The video is shot from the perspective of the driver, providing a comprehensive view of the car's interior. The car's interior is immaculately clean and well-maintained, reflecting the car's high-end status. The video does not contain any text or additional elements, focusing solely on the car's interior. The video's style is straightforward and informative, providing viewers with a detailed look at the car's interior design and features."
+    ],
+    "video_ids": [
+        "zke_154jeBg_23_0to113"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, digital display, control buttons, seats upholstered in black leather? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a steering wheel with the Lamborghini logo, a digital display behind it, multiple control buttons on the center console, and seats upholstered in black material with red stitching, which aligns with the described elements. The presence of additional elements like the car's exterior and background does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a steering wheel, a BMWorghini logo, a digital display on the, and control buttons on the dashboard console, and seats upholstered in black leather with red stitching. which aligns with the description elements. The presence of additional elements like the car's exterior and the does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2b783abddafa8d245517178134527a7bf0c6effc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95267591ac265a2caeef25bd42368b31dad2d2345eeabf45e5393b58bde510d8
+size 271387
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..cdbd39dfdb16f3b45d3d964f5498264a1e473d9d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a meal being served on a green plate. The meal consists of a variety of dishes, including rice, meat, and vegetables. The plate is placed on a table, and there is a bottle of water in the background. The style of the video is a simple, straightforward food presentation, with no additional elements or actions. The focus is solely on the meal and the plate it is served on. The colors in the video are vibrant, with the green of the plate contrasting against the various colors of the food. The lighting is bright, highlighting the textures and details of the meal. The overall impression is one of a delicious, well-presented meal, ready to be enjoyed."
+    ],
+    "video_ids": [
+        "el_DEU0KwcA_1_0to110"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Green plate, rice, meat, vegetables, bottle of water? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a green plate with rice, meat (chicken), vegetables (green beans and carrots), and a bottle of water in the background. All specified objects are present and accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a green plate with rice, meat,whichicken), and (tom and, red), and a bottle of water in the background. The the objects are present, match depicted in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c6c9c51ec4bb5c593e41b5f4d51028e64896a4d2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:53345ba7c19d9230406f4fb61e56a8ae169572e3a9b298704b16fde22d12b4e8
+size 139364
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..cbd69a4cf7d760f545d22eec60f4c05464d87fae
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are standing behind a table filled with various bowls and plates of food. The man is wearing a white chef's coat, while the woman is dressed in a blue shirt. They are both smiling and appear to be presenting the food to the camera. The table is set up outdoors, with a city skyline visible in the background. The food on the table includes a variety of dishes, such as sushi rolls, salads, and bowls of soup. The overall style of the video is casual and inviting, with a focus on the food and the people presenting it."
+    ],
+    "video_ids": [
+        "1Vgpmg1kdCQ_0_129to262"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a chef's uniform and a woman in a blue top standing behind a table, which matches the 'Object(s): A man and a woman' condition. Additional elements like the table setup, background scenery, and other objects do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man and a chef's uniform and a woman in a blue blouse, behind a table filled which ful the descriptionA(s)' A man and a woman' condition. The elements like the table with with bowls,, and the people do not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c035b463a9702f3497af0e773e1b8b6d4debb417
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:527bfc52c766f7bf1ef7dafd12f4ecc26425f15f18b1fae63a0b19c808ab7465
+size 141591
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..094f239ac56e0ffec42e83fad03b783d133fc7ce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a river with a bridge in the background. The first frame shows a close-up of a lifebuoy, which is tied to a metal pole. The lifebuoy is orange and white, and it's floating on the calm water. The second frame shows the same lifebuoy, but now it's further away from the camera, giving a wider view of the river. The third frame shows the bridge in the background, with the lifebuoy still visible in the foreground. The bridge is made of stone and has a metal railing. The river is calm and the sky is clear, making for a peaceful and tranquil scene."
+    ],
+    "video_ids": [
+        "d3NyT0j656A_13_0to129"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Lifebuoy, metal pole, bridge? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a lifebuoy (orange with white rope), a metal pole (part of the boat's structure), and a bridge (large structure spanning the water with stone pillars). These objects are prominently featured and align with the described condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a lifebuoy attachedred and white stripes), a metal pole (supportially the life's structure), and a bridge inin structure in the water). rail walls). The elements are present featured and match with the description '.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e0f3018948f410826aa1a60b34c97fc05c4e1ca1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21c4438a947673d34850494bbcfd52c62400d996d7b0912417172e8007547ae5
+size 264466
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..0a18d218994822c575f067f7efb371f8191bd556
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men engaged in a conversation. The first man is wearing a black jacket and has a serious expression on his face. The second man is wearing glasses and a black sweater, and he appears to be listening intently to the first man. They are standing on a porch with a white railing, and there is a tree in the background. The lighting suggests it is daytime. The style of the video is realistic and it seems to be a scene from a movie or TV show."
+    ],
+    "video_ids": [
+        "Z4zb5uepYw8_28_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men are present. The first man wears a black jacket and has a serious expression. The second man wears glasses and a black sweater, and he is listening intently.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men. The first man is wearing a dark jacket (appears black) and has a serious expression. The second man is wearing glasses and a black sweater, and he appears to be listening intently. These details align with the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men. The first man is wearing a black jacket andlikelyars to in and has a serious expression. The second man is wearing glasses and a black sweater, and he appears to be listening intently. The details match with the description provided.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c98bf8fdcee25693535b8a7a53b108e3c1faa323
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0c5410cb1096e0da8b7c36ccdcef61bf84b5045bc4f8a139d9113f318b19557
+size 161431
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..8a7f8140508d470cea152afccb4b17703b90d0b2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, three women are sitting on a couch in a room with a white wall. The woman on the left is wearing a black tank top and glasses, the woman in the middle has red hair and is wearing a blue shirt, and the woman on the right is wearing a purple scarf. They are engaged in a conversation, with the woman on the right holding out a piece of chocolate to the woman in the middle. The room has posters on the wall, and the overall atmosphere is casual and friendly."
+    ],
+    "video_ids": [
+        "3ymVuFWSv2U_9_90to265"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three women, a piece of chocolate? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows three women sitting on a couch and interacting. They are holding and discussing pieces of chocolate, which fulfills the 'Object(s)' condition. Additional elements like posters on the wall do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows shows three women sitting together a couch. one with One are holding what examining a of chocolate, which alignfills the 'Object(s)' condition. The elements like the on the wall do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c4ddc6d61d63e47dd8c7dfc2807ea83bbca6ee06
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ddf5e5d09e6a6a7b0055afdc45589fef56c69a12ec3661f1d5d2650880996480
+size 140786
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..de12d6cd705e8bcd933d7c068438960932da996c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red pickup truck driving on a snowy road. The truck is equipped with black rims and large tires, suitable for off-road conditions. The truck is moving forward, and the camera angle is from the side, capturing the vehicle's motion. The snow-covered road and surrounding landscape indicate that the video was taken during winter. The style of the video is realistic, with no visible filters or artistic effects. The focus is on the truck and its movement through the snowy environment."
+    ],
+    "video_ids": [
+        "MFVxvRxpyjw_52_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red pickup truck? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red pickup truck, specifically a Toyota Tacoma TRD 4x4 Off-Road model, as described. The truck is the central focus of the video, and its red color and pickup truck design are consistent with the 'Object(s)' condition. Additional elements like snow and trees do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a red pickup truck, which a Ford T,D Pro4x4, Roadoad,, which indicated. The truck is prominently central focus of the video, and it red color and off truck design are prominently with the 'Object(s)' condition. The elements like the and a in not contradict the core but}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4b473f7425355091104d9fa998259cc82c2fb62f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:050cf66b7272656085c0e0fc501dbfa5745b5e6a1101475eac0abd68a3534508
+size 178906
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..435452f8d24b4f32b0c5208776436f92b296e275
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene lakeside scene during what appears to be either sunrise or sunset, as indicated by the warm hues of orange and yellow reflected on the water's surface. The sky is a mix of soft clouds and streaks of light, creating a tranquil atmosphere. On the left side of the frame, a line of trees stands tall, their silhouettes adding depth to the landscape. In the foreground, a duck is seen swimming gracefully across the calm lake, its reflection mirroring its movements on the water's surface. As the video progresses, another duck enters from the left side of the frame, moving towards the center. The ducks' actions are slow and deliberate, contributing to the peaceful ambiance of the scene. The overall setting remains consistent throughout the video, emphasizing the natural beauty and tranquility of the environment."
+    ],
+    "video_ids": [
+        "a23e73157f848babc84e7b1940e47b1180216810d9f1c7f01fc94c5fc8f1b687"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Trees (left side), Ducks (swimming and moving towards center)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows trees on the left side of the frame, and ducks are visible swimming and moving toward the center. The ducks' movements and the presence of trees align with the described condition, even though the sky has a sunset hue and there is a distant smokestack, which do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a on the left side of the frame, which there are swimming swimming and moving towards the center of The presence are movement and the presence of trees align with the description conditions, making though the trees and a sunset-like which the are a reflection mountain-like, which are not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6705fa3f6650a4e9dcebf16e4a43696ca229c7e9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1125-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5210a911c545d55978c8be4ea3d8448e8b6b17037b28774c855443844ede743f
+size 175472
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..09d14f9e7b0ed73b2e871b031e2cf0bd635dea4a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young girl with blonde hair and a pink bow in her hair. She is wearing a pink dress with a floral pattern. The girl is standing on a wooden deck with a railing. The deck is located in a forested area with tall trees. The girl is looking off to the side, and her expression is one of concern or worry. The lighting in the scene is dim, suggesting that it might be evening or early morning. The overall style of the video is realistic with a touch of animation, giving it a unique and charming look."
+    ],
+    "video_ids": [
+        "-Tsy9mQDXbE_46_20to215"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young girl with blonde hair and a pink bow, wearing a pink dress with a floral pattern.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young animated girl with blonde hair styled in pigtails, each adorned with a pink bow. She is wearing a pink dress with a visible floral pattern, matching the description. The background and setting do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a young girl girl with blonde hair, with aigtails, wearing adorned with a pink bow. She is wearing a pink dress with a floral floral pattern. which the description provided The setting appears her are not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..26657e968012cda372a6f8a9b139f8762a395c70
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5107b317ecebab9547093f884c5cfcec30ce7169a9e86541d1a68662c56b932
+size 137761
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..c6081d7ebf33829a61f83ea96baeb1da5a29c17e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a small brown puppy in a grassy field. The puppy is seen walking through the tall grass, its fur slightly matted from the outdoor elements. The puppy's eyes are wide and curious, taking in the surroundings. The background is a blur of green and red, suggesting a mix of trees and flowers. The puppy's ears are perked up, indicating alertness. The overall style of the video is a close-up, focusing on the puppy's face and upper body. The camera angle is slightly above the puppy, giving a sense of the puppy's height in relation to the grass. The video is a snapshot of a puppy's exploration of the outdoors."
+    ],
+    "video_ids": [
+        "9F7CAXiltLo_3_0to139"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small brown puppy? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a small brown puppy as the main subject, which matches the 'Object(s)' condition. Additional animated elements (like the red stick figure and cartoon dog) do not contradict the core description of the puppy being present and central to the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a small brown puppy standing the main subject. which align the descriptionObject(s)' condition. The elements grass,like the grass flowers and) the characters) are not contradict the core description of the puppy. the.\"\n are to the video.\"\n}<|im_end|>\n\ufffd\ufffd",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cbb215cb156ad4bf9e79d6e747791f845e3432d3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d1fd3eae3912f65a5eea2a75775c9f65ba54e92bbd152f2d67bb8e053fd1d78
+size 119642
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..5aa06a18cbcc3302780e6e0d2e2d901dbc5c8787
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a pink plastic object resembling a bear's head, with a black circular opening at its center. Inside the opening, there is a small white object, possibly a marshmallow or a similar item, which is being manipulated by an unseen hand. The hand appears to be pressing down on the white object, causing it to compress slightly and release some white substance, which spreads out around the red flower-shaped object inside the black bowl. The background is a textured surface, likely a tablecloth or fabric, providing a neutral backdrop for the action taking place. The camera remains stationary throughout the sequence, focusing tightly on the interaction between the hand, the white object, and the red flower-shaped item within the black bowl. There are no other significant changes or movements in the scene, maintaining a consistent focus on the central activity."
+    ],
+    "video_ids": [
+        "f504453937f707e0bca74b932813ed64f93a3a9bbfcebd1348bb014fb19e4333"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A pink plastic bear's head with a black circular opening, a small white object inside the opening, a red flower-shaped object inside the black bowl, and an unseen hand.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a pink plastic bear's head with a black circular opening, containing a small white object and a red flower-shaped object inside the black bowl. The description of an unseen hand is not visually confirmed, but since the core objects are present and no contradictions exist, the video largely fulfills the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a pink plastic bear's head with a black circular opening. a a small white object. a red flower-shaped object inside a black bowl. An unseen also the unseen hand is fulfilled directly confirmed, but the the hand elements match present and the conflicting are, the video largely fulfills the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cd868172363224a84b517fd94f4bcb62d4374d92
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a54d4a465e279c0af6a5a47853ae7b84217c3e94d4fc1b6b43920bcdd35ffec3
+size 116915
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..1a346b3108768973abf4ec83387e50d5abd98bed
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the launch of a rocket from a SpaceX facility. The rocket, with its bright flame and smoke, ascends into the sky, leaving behind a trail of smoke. The rocket is surrounded by a large cloud of smoke and steam, indicating the power of the launch. The rocket is visible in the sky, moving upwards. The SpaceX facility is visible in the background, with a large building and a water tower. The video is a dynamic and exciting depiction of space exploration and technology."
+    ],
+    "video_ids": [
+        "X3faHvP6oeM_8_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A rocket, a large cloud of smoke and steam, a large building, and a water tower.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a rocket launching, producing a large cloud of smoke and steam. A large building with 'SPACEX' branding is visible, along with a water tower in the background. All core elements described are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a rocket launching, with a large cloud of smoke and steam. There large building is aNASAACEX' written is visible in and with a water tower in the background. The elements elements of in present, match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fe092d682d4048b49a7261ab4c40dfee8c9f58d1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8fe30e32f505c605bf630a52ef60927c6c37346401bf0377bdf0b5a4a37d883a
+size 122948
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..66b887b4a4854b6ab08140c28feab8cd3a3ce298
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene sunset scene featuring a large tree silhouetted against a vibrant orange sky. The sun is positioned low on the horizon, casting a warm glow that illuminates the branches and leaves of the tree. Perched on the branches are several birds, their dark forms contrasting sharply with the bright background. As the video progresses, one bird takes flight, adding a dynamic element to the otherwise tranquil setting. The background reveals a series of rolling hills or mountains, their outlines softened by the distance and the warm light of the sunset. The overall atmosphere is peaceful and picturesque, emphasizing the natural beauty of the moment."
+    ],
+    "video_ids": [
+        "06f2bd1ea4941d45781c5844023aad0213054af156ff91fe4f5643e00282dd27"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large tree, several birds perched on its branches, and rolling hills or mountains in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a large tree with silhouetted branches and leaves, several birds perched on the branches, and rolling hills or mountains in the background, all set against a sunset. The core elements described are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a large tree with severalhouetted branches and leaves, several birds perched on its branches, and rolling hills or mountains in the background. all of against a sunset sky The presence elements of in present represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cd1abc1a2de39a1fb12d7f6b8a00f5977c2a1c2d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a25217432792c178f9857319454d840237de8dfaff225425c8bf71ae6f38ef47
+size 80369
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..839ad44f8a7b26f12d5292059a146024b6f9b595
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a blue and gray hoodie, standing in a locker room with a blue wall in the background. He appears to be deep in thought, looking down and to the side. The man's expression is serious, and he seems to be contemplating something. The locker room setting suggests that he might be a professional athlete, possibly a football player, given the context of the image. The overall style of the video is candid and unposed, capturing a moment of introspection in the athlete's life."
+    ],
+    "video_ids": [
+        "5OkPr5pOke8_5_0to154"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue and gray hoodie, appearing to be deep in thought, looking down and to the side with a serious expression.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue and gray hoodie, and his facial expression and gaze suggest he is deep in thought, looking down and to the side. The background and attire are consistent with the description, and no elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue and gray hoodie, which he posture expression and posture direction he is deep in thought, looking down and to the side with The serious is lighting match consistent with the description, and there additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..83cfcc8e30e2b1fa09f0fe83b1aea54c883a1bce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d45a4bc4843dec7f20867f1b031df39441bce95a85aa71e2eedafbae7c5e8b1f
+size 127578
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..393da95d59a7224a724d06e5d59068268f064821
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a white guitar pedal with multiple knobs and a label that reads \"MAKO D1\". The pedal is connected to a black guitar pedal with a label that reads \"E\". The guitar pedal is placed on a wooden surface. The style of the video is a product demonstration, showcasing the features of the guitar pedal. The focus is on the pedal and its connections, with no additional context or background provided. The video is likely intended for an audience interested in music equipment or guitar pedals."
+    ],
+    "video_ids": [
+        "ldvHR5IyL_4_3_0to185"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: White MAKO D1 guitar pedal with multiple knobs, Black guitar pedal labeled 'E', Both guitar pedals.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a white MAKO D1 guitar pedal with multiple knobs and footswitches, as described. Additionally, a black guitar pedal labeled 'E' is visible to the right, confirming the presence of both pedals. The video matches the core description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a white MAKO D1 guitar pedal with multiple knobs, aswitches, which well. Additionally, there black guitar pedal labeled 'E' is also in the left of which the presence of both guitar as The video does the description description of any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3c48d55c5c37b0b97628ecbe786489828adae557
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de775704178777b31736fa9f3410cdde6d804f8eac43ba2b4b7a9b7a4a0ccf27
+size 83349
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..a4de6c4c8371dc743f9071d7505a1577cb3e2af1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are seen walking together in a park-like setting. The man is wearing glasses and a blue shirt, while the woman is dressed in a red and white striped shirt and sunglasses. They are both smiling and appear to be enjoying their time outdoors. The park is lush with green grass and trees, and there is a building visible in the background. The overall style of the video is casual and relaxed, capturing a pleasant moment between the two individuals."
+    ],
+    "video_ids": [
+        "-acPjaRI3O0_1_0to172"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman walking together. The man is wearing glasses and a blue shirt. The woman is dressed in a red and white striped shirt and sunglasses. Both are smiling.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing glasses and a blue shirt, and a woman in a red and white striped shirt with sunglasses. Both are smiling and appear to be standing together outdoors, which aligns with the description. Although they are not explicitly walking (they are standing still), the core elements of their appearance and expressions match the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and glasses and a blue shirt, and a woman wearing a red and white striped shirt and sunglasses. Both are smiling and appear to be walking close in, which aligns with the description of The they are not walking shown,as are standing),), the context elements of the appearance and the match the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..37b5f90302fd88716450e47b342828864823ef4d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1126-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2eefdb9df2751baee0cd393588f44b2ad5258f28487f12467766f93215341ce2
+size 332445
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..1b5a819cb4fe005ae1e9218e625f3e0d92277d85
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a large, gray Lego shark sculpture with a red tongue, standing on a blue Lego base. The shark is positioned in front of a white building with a blue roof and a palm tree. The background is filled with lush green trees and a clear blue sky. The style of the video is a time-lapse, capturing the Lego shark sculpture being built from start to finish. The camera angle is from a low perspective, looking up at the sculpture, emphasizing its size and detail. The overall atmosphere of the video is playful and creative, showcasing the impressive craftsmanship of the Lego sculpture."
+    ],
+    "video_ids": [
+        "6UEAKx9Itjg_2_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Lego shark sculpture, blue Lego base, white building, blue roof, palm tree, lush green trees, clear blue sky? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a large Lego shark sculpture with a blue Lego base, set against a backdrop of a white building, blue roof, palm trees, lush green trees, and a clear blue sky. All elements mentioned in the condition are present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a Lego Lego shark sculpture with its blue Lego base, positioned in a backdrop of a white building with a roof, palm tree, lush green trees, and a clear blue sky. The the described in the description are present and match depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ca624a5e05a27182d52834c769532ae328cbff83
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:27c69fe41ba207dad606256b1830de71dbecb1709ebf8d2f3e10878b1e1d4fbd
+size 108646
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..774d19d5c1659569affaf4e66a8ef6a7fbaf9b3e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a hat, who appears to be in a jungle setting. He is holding a small, green frog in his hand. The man's facial expression is one of surprise and excitement, as he looks at the frog with wide eyes. The background of the video shows lush greenery, including ferns and other plants, which suggests that the man is in a tropical or subtropical environment. The style of the video is realistic, with a focus on capturing the man's reaction to the frog and the natural surroundings. The video likely has a documentary or educational purpose, as it provides a glimpse into the man's experience with the frog and the environment in which it lives."
+    ],
+    "video_ids": [
+        "SsgGBYTVkXw_34_0to119"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a hat, holding a small, green frog in his hand.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a hat, holding a small green frog in his hand. The frog has distinct black markings, and the man appears to be in a natural, green, leafy environment. All elements align with the described condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a hat, holding a small, frog in his hand. The frog is a green eyes and which the man appears to be outdoors a natural, outdoor environment leafy environment, The elements of with the description '.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cfbe9d6cbc404ff715eb673126ef76b331dc4c24
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c03118275953b5a2148da41d051837b9b8ed541a4db69c16be469728a872a8fe
+size 137444
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..7502637832dc8696cc94f91daa19c999700ed38f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two men are sitting in a room with a Christmas tree in the background. The man on the left is holding a cup, while the man on the right is holding a glass of beer. They are engaged in a conversation, with the man on the right gesturing towards the man on the left. The room is decorated with colorful balloons and a banner that reads \"BABY BEATS\". The overall atmosphere of the video is festive and casual."
+    ],
+    "video_ids": [
+        "iuq2qWEXZ4A_9_435to584"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, a cup, a glass of beer, a Christmas tree, balloons, and a banner.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men, one holding a cup and the other holding a glass of beer. A decorated Christmas tree is visible in the background, along with colorful balloons and a banner that reads 'ABIEATS'. All specified objects are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two men sitting one holding a cup and the other a a glass of beer. There Christmas Christmas tree is visible in the background, along with balloons balloons and a banner. appears 'MBA'.'. The the elements are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..922b2060bebdc6cc9ddd1cb0ea1ef4c32d7eb83b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13db5eb26cc5d9d9715ac6f3b9c248bd7b6bb2cd78f5d155cb50ed0756018a97
+size 141853
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..ea408f5af2bd8cbcb81223dfa5b0777ce55e687f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit sitting in a chair, looking thoughtful. The setting appears to be a well-lit room with a lamp in the background. The man is wearing glasses and a tie, suggesting a formal or professional context. The video is likely a news segment or an interview, as indicated by the text overlay that reads \"Countering the 'mise' canard.\" The style of the video is straightforward and professional, with a focus on the man's expression and the message conveyed by the text."
+    ],
+    "video_ids": [
+        "Hs74tp7DYbM_55_0to153"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a suit, glasses, tie, sitting in a chair; text overlay reading 'Countering the 'mise' canard'.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a suit, glasses, and tie, sitting in a chair, which matches the core description. Additionally, a text overlay reading 'Countering the 'mise' canard' is visible, fulfilling that part of the condition. The presence of a lamp and framed artwork in the background does not contradict the description and is acceptable as an additional element.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit, glasses, and a, sitting in a chair, which matches the description description. The, there text overlay reading 'Countering the 'mise' canard' is present, which the part of the description. The presence of the lamp in a picture in the background does not contradict the description and is acceptable as additional additional element.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0fbcee718631dbaa44aa7ad4b4b525bf6fa62b3b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e15f4f8ff422d371b967238f091743d5f2a651d61598263723dc474b87d4c2a
+size 98374
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..bf8e7241d30950e6620b9afa65002ef08b84eccd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are sitting on a couch, engaged in a conversation. The man is wearing a black jacket and a white shirt, while the woman is dressed in a black tank top. They are seated in a room with a mirror on the wall behind them. The room also features a red couch and a striped pillow. The woman is holding a brown purse. The video captures a casual and intimate moment between the two individuals."
+    ],
+    "video_ids": [
+        "ABw-PkNbrgE_3_0to164"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a woman, a black jacket, a white shirt, a black tank top, a brown purse? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black jacket and a white shirt, and a woman wearing a black tank top. The woman also has a brown purse strap visible over her shoulder. These elements are clearly present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a black jacket over a white shirt, and a woman wearing a black tank top. There man is has a brown purse on visible, her shoulder. The elements match consistent present in match the description.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..91cd9e315a75ef2def0d141e1488db70030672e4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:10d0671f8efe51f8886e0d4f76f54218583c776d324c638079b6ca231b0bfe73
+size 92016
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..0eedddb73c86250a96034f797051302cffc4a0fa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene outdoor scene featuring a small bird perched on a tripod-mounted camera. The bird, likely a titmouse given its size and coloration, is predominantly blue and white with a yellow underside. It is positioned near the camera lens, facing slightly to the right. The background is softly blurred, showcasing a mix of green foliage and hints of a wooden structure, possibly a birdhouse or part of a fence, suggesting a garden or park setting.  Initially, the bird remains stationary, observing its surroundings with keen interest. As the video progresses, the bird begins to move, shifting its position on the tripod. It turns its head from side to side, appearing curious about something off-camera. The bird's movements are gentle and deliberate, indicating a calm and undisturbed environment. The camera itself remains fixed throughout, maintaining a steady focus on the bird and the camera setup, allowing viewers to observe the bird's natural behavior in detail. The overall atmosphere is peaceful, highlighting the interaction between"
+    ],
+    "video_ids": [
+        "47753107c29389ea52d42a43fff0d7480ca2b314f82f05e96f68ebc581e8f571"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small bird (likely a titmouse) perched on a tripod-mounted camera.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small bird, consistent with a titmouse (blue tit), perched on the tripod-mounted camera. The bird is clearly visible, interacting with the camera, and the setup matches the description. Background elements like branches and foliage do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small bird, which with a titmouse,likely tit), perched on a top-mounted camera. The bird's positioned visible and and with the camera, which the setting matches the description provided The elements like green and green are not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..637c0ba24c5b970281955d6473136ae1884d761b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7725127a5147a4f19430744e54f36f4f2b09c474da5bb77220827d7b04171f0d
+size 151881
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..f3e95772410c90295a7fa8a8d33184ebc7183dae
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a journey on a winding mountain road. The road, painted with a yellow center line, curves through a rocky landscape, with steep cliffs on either side. The road is flanked by sparse vegetation, including small trees and shrubs. The sky above is a clear blue, with a few clouds scattered across it. The sun is shining brightly, casting a warm glow on the scene. The perspective of the video is from a high angle, providing a bird's eye view of the road and the surrounding landscape. The video is a testament to the beauty of nature and the thrill of driving on a mountain road."
+    ],
+    "video_ids": [
+        "-VrQvDjZXSk_29_0to164"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Winding mountain road, yellow center line, rocky landscape, steep cliffs, small trees and shrubs, and a clear blue sky with scattered clouds.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a winding mountain road with a yellow center line, surrounded by a rocky landscape with steep cliffs, small trees and shrubs, and a clear blue sky with scattered clouds. All described elements are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depictsfills the 'Object(s)' condition as depicting depicting a winding mountain road with a yellow center line. a by a rocky landscape with steep cliffs. small trees and shrubs, and a clear blue sky with scattered clouds. The these elements are present and contribute with the scene.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..27f8dd9630b29cd1e71e226b4783bc7fdb847a19
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9d95dc918a46f191ace5f92e8e4c1ab46965c85c7a3712d31ef27dccea9bd569
+size 258937
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..2c66e35834ef318150140b715e2082252eb8373c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man sitting at a table, looking at his cell phone. He is wearing a black shirt and a black baseball cap. The man is holding the phone in his hands, and he appears to be focused on the screen. The table is white, and there is a gray background behind the man. The man's expression is serious, and he seems to be concentrating on whatever he is viewing on his phone. The overall style of the video is simple and straightforward, with a focus on the man and his interaction with the phone."
+    ],
+    "video_ids": [
+        "GS2Ef8fFCTw_28_23to187"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a cell phone, a black shirt, and a black baseball cap.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a black shirt and a black baseball cap, holding and interacting with a cell phone. All specified objects are present and accurately represented in the video without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing a black shirt and a black baseball cap, holding a looking with a cell phone. The the elements are present and match depicted in the video.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0e88541de2e263e3ade8e3906789928f0eb3484c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1127-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:18b0e4ccadbf1d75359d7272e350b42ebe9db9a60a6398991e25bbe710552562
+size 65757
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..75a0c4804c9d01ae51dfb41bfa499d9d6b7593f1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen preparing a unique sandwich on a wooden cutting board. The sandwich is made with a sesame seed bun, which is filled with a generous amount of chicken, melted cheese, and a surprising addition of orange cheese puffs. The person's hands are visible as they assemble the sandwich, adding each ingredient with care. The overall style of the video suggests a casual and fun approach to food preparation, with a focus on creating a visually appealing and delicious meal."
+    ],
+    "video_ids": [
+        "V47VTKKveP0_47_119to310"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person, wooden cutting board, sesame seed bun, chicken, melted cheese, orange cheese puffs? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a person cutting a burger on a wooden cutting board. The burger features a sesame seed bun, chicken, melted cheese, and orange cheese puffs, all of which are clearly visible and match the description. No contradictory elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a person interacting into sesame on a wooden cutting board. The burger contains a sesame seed bun, a, melted cheese, and orange cheese puffs. which of which are present visible and match the description. The additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8b0bb14e87f3146808babfcbf79bdb5ddaf4bab3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:907a0887049cf768616d07b7a3b34c80d395729920379f2fcbdf66c89278b023
+size 212916
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..d7430f04b1d80955ba1f1a09643e990812f75bc5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated depiction of a vintage fire truck on a track, with two animated firemen in blue uniforms and yellow helmets. The fire truck is red with a gold grill and a large bell on top. The firemen are standing on either side of the truck, one on the left and one on the right. The truck is on a track that is surrounded by grass and rocks. In the background, there is a brick building with a large door and a window. The style of the video is cartoonish and colorful, with a focus on the fire truck and the firemen. The animation is smooth and the colors are vibrant. The video does not contain any text."
+    ],
+    "video_ids": [
+        "cy14sJz-ltg_36_0to145"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A vintage red fire truck, two firemen in blue uniforms and yellow helmets? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a vintage red fire truck with a hose reel on top, consistent with the description. Two firemen in blue uniforms and yellow helmets are visible; one is driving and the other is seated beside him. The scene also includes a brick building and a railway track, which do not contradict the core description and are acceptable as additional elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a vintage red fire truck with a classic attached on the, which with the description. There charactersmen in blue uniforms and yellow helmets are standing, one is standing the the other is standing beside the. The setting is includes a building building and a grass track, which are not contradict the core description but are additional additional additional elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..facb18ce0282fc13d9ff6cb256ee20f9acb81727
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:96d4ccf1eac893ed82746000516e8fa616595362441ff0a426bcdd9cbb986eeb
+size 160897
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..e3ff78041be86b59a810fbe765e80a168131fefc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of camaraderie between two men, both dressed in blue shirts, sitting side by side in a stadium. The man on the left, with a serious expression, is looking off to the side, perhaps lost in thought or observing something in the distance. His companion on the right, however, is smiling broadly, his eyes sparkling with joy. The background is filled with the blurred figures of other spectators, their faces indistinct but their presence adding to the atmosphere of the event. The focus of the video is clearly on the two men, their contrasting expressions and the shared experience they are having."
+    ],
+    "video_ids": [
+        "AudG5lTDyww_0_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men in blue shirts? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features two men wearing blue shirts, which matches the 'Object(s)' condition. Although there is a woman in the background, her presence does not contradict the core description, as the condition only requires the presence of two men in blue shirts, which is clearly fulfilled.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features two men wearing blue shirts, which align the 'Object(s)' condition. The there are a crowd in the background, the presence does not contradict the main description of as the focus only specifies the presence of two men in blue shirts.\"\n not is clearly fulfilled.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..776b177b831cc952efb25a3514c6e1b96df349bb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54a3e016fd87a15089033d788219bc5f1f56a774e9ac4cbdd4c6cf6372e6ef46
+size 206564
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..61e23f22b9e0dbfdaac75f8dfe2739b7a05df0e3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a car's engine compartment, focusing on the engine cover. The engine cover is silver and black, with a distinctive logo on it. The engine cover is opened, revealing the engine underneath. The engine is complex, with various components and hoses visible. The video is likely educational, providing a detailed look at the engine's components and layout. The style of the video is informative, with a focus on the technical aspects of the engine. The video is likely aimed at car enthusiasts or mechanics, providing a detailed look at the inner workings of the engine."
+    ],
+    "video_ids": [
+        "PtiYKdxL4OA_27_0to178"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Engine cover, engine? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the engine cover and engine components of a car, with the engine cover prominently displayed and detailed. The description matches the visual content without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a engine cover being engine of of a car. which the cover cover being displayed in then. The sequence of the content content of any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2995bab8e93011aaf768b909a90a78c195ac8a33
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5aa128a916e73f0753bf593fc71fc2dbc477fb06e450fa2704edee338d9a0be0
+size 249820
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..4a0ad518593a44485cd2e0dc5fe971356d1e63fa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a fish being prepared for cooking. The fish is placed on a yellow plate, which is placed on a countertop. The fish is covered in a red sauce, and it appears to be seasoned. The fish is the main focus of the video, and it is shown in close-up. The countertop is made of concrete, and there is a bottle of sauce in the background. The style of the video is realistic, and it captures the process of preparing the fish in detail."
+    ],
+    "video_ids": [
+        "2WqIQ7nel0k_27_0to134"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A fish, a bottle of sauce? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a fish on a plate, covered in seasoning, and a bottle of sauce is visible in the background. These elements match the 'Object(s)' condition described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a fish covered a yellow, covered in a or which there bottle of sauce in visible in the background. The elements match the descriptionObject(s)' condition described in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..141f24c3297756f6358f949abdd79bb9fb06f680
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63c7a966aea56a39cf324ad15ef5afddc5c244f50b225e680ea8d184cc72dff3
+size 126886
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..a766e27499b5457c28c837080f262683614946cc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen in a room with a white wall and a pink shelf in the background. He is wearing glasses and a blue shirt with a vest. In the first frame, he is looking at a toy robot with a face, which is being held by someone off-camera. In the second frame, he is still looking at the toy robot, which is now being held by him. In the third frame, he is still holding the toy robot and looking at it. The toy robot has a face and is white in color. The man appears to be intrigued by the toy robot. The room has a casual and relaxed atmosphere."
+    ],
+    "video_ids": [
+        "_uLmn_Dba08_45_25to146"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a toy robot with a face, a blue shirt with a vest, a white toy robot, and an off-camera person holding the toy robot initially.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue shirt and a vest, holding a white toy robot with a face (Olaf from Frozen). An off-camera person's hand is visible initially, holding the top of the toy robot. The core elements described are all present and consistent with the video content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man interacting a blue shirt with a vest, holding a toy toy robot with a face.whichaf from Frozen). There off-camera person is hand is visible,, holding the toy of the toy robot. The man elements of in present present and match with the video content.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e3e4eeeb15efbea9850e2f9d550191d080727567
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec4a41e53608ad650e408ef3aa137b0fd62ad9ca2ed123fa2ec40f811b389e98
+size 201266
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..5c3f27288f037fc1772337a22296fed0936ac555
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red sports car parked in a garage. The car is sleek and shiny, with a black stripe running down the side. The car's tail lights are on, adding a touch of color to the scene. The garage is spacious and well-lit, with a concrete floor that reflects the car's red color. The car is parked in the center of the garage, drawing attention to its sleek design and vibrant color. The overall style of the video is realistic, capturing the details of the car and the garage with precision. The video does not contain any text or additional objects. The focus is solely on the red sports car and its surroundings."
+    ],
+    "video_ids": [
+        "-K28RLGsTYg_18_88to212"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red sports car with a black stripe running down the side, with its tail lights on.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red sports car with a black stripe running down the side, and the tail lights are visibly illuminated. These elements match the description provided, even though the video also includes additional elements like a 'Please Stand By' text overlay and background details, which do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red sports car with a black stripe running down the side, which the tail lights are illuminated illuminated. The elements match the description provided, indicating though the video is includes a elements like the spepol do By' message and, a noise that which do not contradict the core description of}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7588ec4fc7029a845d647d474b7480c2da4dc14d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c560fbac14a78784a0d19e2886549e1874dab27b18ddb5b18df95d31239ac85
+size 201961
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..8fc16053adb704d9118aa8c224512c8454bd2bc8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a delightful tea party set up on a table. The table is adorned with a variety of tea cups, each with a unique design, and saucers. The cups and saucers are arranged in a visually pleasing manner, creating an inviting atmosphere.   In addition to the tea cups and saucers, there are also wine glasses and bottles of wine on the table. The wine glasses are elegantly shaped and filled with wine, ready to be enjoyed. The bottles of wine are neatly arranged, adding to the overall aesthetic of the table setting.  The table is set against a backdrop of a lush green garden, which adds a touch of nature to the scene. The garden provides a serene and peaceful setting for the tea party, making it an ideal place for a relaxing afternoon.  Overall, the video presents a beautifully set tea party, complete with tea cups, saucers, wine glasses, and wine bottles, all arranged on a table in a picturesque garden setting. The attention to detail in the table setting and the tranquil garden backdrop create a sense of elegance and sophistication."
+    ],
+    "video_ids": [
+        "zI5i4gEHEqY_1_0to103"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Tea cups, saucers, wine glasses, bottles of wine? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows multiple tea cups with saucers, wine glasses, and bottles of wine (both green and pink bottles labeled 'bottle green'). These items are prominently displayed on the table, fulfilling the 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows tea tea cups and saucers, wine glasses, and bottles of wine.both red and dark labels). aswineott'' The objects are the displayed on a table, fulfilling the 'Object(s)' condition. any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ff69e8b153e4624bec93129df586def1833a664a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1128-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e5c7ace30a70af109c858c4589a75825ce4042bd600c31dd5b580b11454dbb2
+size 165961
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..a86404a34e6856c223b0042cb049ce5754b65c5f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a colorful and festive celebration of a birthday party. The main focus is on three bottles of soda, each wrapped in a colorful ribbon and adorned with a label that reads \"Happy Birthday Ellie\". The bottles are placed on a table covered with a vibrant yellow tablecloth, which is scattered with confetti and party decorations. In the background, there's a cake with a colorful striped design, adding to the festive atmosphere. The style of the video is cheerful and lively, capturing the essence of a joyful birthday celebration."
+    ],
+    "video_ids": [
+        "s1te6HHLf2U_15_0to119"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three bottles of soda, a vibrant yellow tablecloth, confetti, party decorations, a colorful striped cake.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing three bottles of soda with personalized labels and straws, a vibrant yellow tablecloth covered in confetti, various party decorations including a colorful striped cake in the background, and other festive elements like candy buckets and cards. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showing showing three bottles of soda, colorful labels, aaws, a vibrant yellow tablecloth, with confetti, and party decorations, a colorful striped cake, the background, and additional party elements like stream and and stream. The these elements are in present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..975d1ad935594de71c4364770bee0da93ef4bbfb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a7e2a710920b40ea5f2a9153dc0603fd8f1f1272885e6275e2bc7848601de471
+size 187118
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..2c0403e88d94159bb7e41552aefbf85a73ebdf10
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a vibrant and colorful bird perched on a branch in a lush, green forest. The bird, with its striking blue and green feathers, is the focal point of the video. It is seen in three different positions on the branch, showcasing its agility and grace. The forest around the bird is dense with foliage, providing a natural and serene backdrop. The bird's movements and the surrounding environment create a sense of tranquility and harmony with nature. The video is a beautiful representation of wildlife in its natural habitat."
+    ],
+    "video_ids": [
+        "F0ppCaEiU2U_55_37to187"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A vibrant, colorful bird with striking blue and green feathers.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bird with vibrant, colorful feathers that are predominantly blue and green, matching the description. The bird is perched on a moss-covered branch in a natural setting, and while there are other elements like branches and background foliage, they do not contradict the core description of the bird's appearance.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a bird with vibrant blue colorful feathers that include predominantly blue and green, which the description provided The bird is perched on a branchy branch, a lush setting, which the there are other elements like green and leaves foliage, they do not contradict the core description of the bird's appearance.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fc4b3db9e2d7aee3d29e6816eb50730d50679fb3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41be14f84ed8f4ea3d732eeb1ef87f35feae9de9bedcb10bf544a7c6099f4f9e
+size 168389
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..6f949c773f6311132c7a1ca59f5f5303d8e31a1d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a sea turtle swimming through a coral reef. The turtle, with its distinctive patterned shell, is the main focus of the video. It is surrounded by several small fish, which are scattered around the turtle as it moves through the water. The coral reef, with its vibrant colors and intricate patterns, provides a stunning backdrop for the turtle's journey. The video is shot from a perspective that follows the turtle, giving the viewer a sense of being right there in the water with the turtle and the fish. The overall style of the video is one of natural beauty and tranquility, capturing the grace and elegance of the sea turtle in its natural habitat."
+    ],
+    "video_ids": [
+        "MVKRBEZK2Co_33_0to131"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A sea turtle and several small fish? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a sea turtle swimming underwater, accompanied by several small fish. The turtle is the central subject, and the fish are visible around it, matching the description. The background coral reef does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows depicts a sea turtle, among, which by several small fish. The turtle is the central focus, and the fish are visible in it, fulfilling the description of The presence consists and adds not contradict the presence description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fcbd693dd6e141dc20cbabfb537f190b5c7f5ba6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b3b65cc641a8672f3fbbd45279a1dab736123708a56a4559a01297bb3d8924f
+size 311132
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..7ee20c253fa19629dc52154022d6e3d75353bbf3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit sitting at a table with a microphone in front of him. He appears to be in the middle of a speech or interview, as he is looking off to the side with a thoughtful expression. The man is wearing a dark suit and tie, and his hands are clasped together on the table. The background is blurred, but it seems to be an indoor setting with a blue wall. The style of the video is a straightforward, professional interview or speech, with a focus on the man and his words."
+    ],
+    "video_ids": [
+        "PQLnmjPaCvM_22_0to153"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a dark suit and tie, sitting at a table with his hands clasped together and a microphone in front of him.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a dark suit and tie, sitting at a table with a microphone in front of him. His hands are clasped together under his chin, which matches the description. The background and other elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a dark suit and tie, sitting at a table with his microphone in front of him. His hands are clasped together, his chin, which align the description. There setting and setting elements in not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..70d8c72eb9423c767fe34aae911331f12e4b55bc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c8fbdc98d53737bbc168009946b77a3afdb7a2145348b2d37ca2dc43e9e5ec6
+size 101782
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..0db41a5d45b03e02dba18f74a66d510eeb44c0a8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a chef in a professional kitchen, showcasing his culinary skills. The chef, dressed in a black apron and a matching black cap, stands confidently behind a wooden counter. In front of him, a white bowl filled with golden-brown fried food, possibly chicken, is placed. The chef holds a spoon, ready to serve or taste the dish. The kitchen is well-equipped with various appliances, including an oven and a sink, indicating a professional setting. The overall style of the video suggests a cooking show or a culinary tutorial, with the chef as the main subject, demonstrating his expertise in preparing delicious meals."
+    ],
+    "video_ids": [
+        "MCpJXQy0bj8_14_55to271"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chef, white bowl with fried food, spoon? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a chef standing in a kitchen, a white bowl filled with fried food (appearing to be fritters or falafel), and a spoon placed in a small compartment of the bowl. These elements match the 'Object(s)' condition described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a chef in in a kitchen, wearing white bowl filled with what food,likelyaring to be donritters or similarafel), and a spoon being in the bowl bowl on the bowl. The elements match the descriptionObject(s)' condition provided in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2c8cef87bd072337d66e8ce865e4f67cb1c379a3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:29c15ca74964e07760dc8821bd770e0b8c79f30d29555b615be4de44e53454e6
+size 118169
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..37030122f4439aa3c0b0bba08a842f5712d1a8af
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a close-up view of a motorcycle in three different stages of assembly. The motorcycle is predominantly black and orange, with a sleek design and a powerful engine. In the first frame, the motorcycle is shown with its engine and frame fully assembled, but without any body panels or components. The second frame shows the motorcycle with the body panels and components partially installed, revealing the intricate details of the engine and frame. The third frame shows the motorcycle fully assembled, with all body panels and components in place, showcasing its complete design and functionality. The style of the video is a time-lapse or stop-motion animation, capturing the transformation of the motorcycle from its bare bones to its final, fully assembled form."
+    ],
+    "video_ids": [
+        "tRw8REZD9so_9_0to148"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A motorcycle? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a motorcycle with visible components such as the engine, exhaust, seat, wheels, and bodywork. The motorcycle is the central and only object in the frame, fulfilling the 'Object(s): A motorcycle' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video consists depicts a motorcycle in a features such as wheels seat, wheels, wheels, and, and handle.. The design is depicted central focus only object in the video, and the 'Object(s)' A motorcycle' condition. any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..108fa7acf703e918b190f2e64b149a8832ebcbaa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f4f07f2312906b8916720d55ddaac6e8ed328b43b81b2f12a529a87bd74ed5f3
+size 148719
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..bcfc6a2648c59bb6778fbf4016d6b996341aa591
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a young girl in a kitchen, dressed in a pink chef's hat and apron, preparing to cook. She is standing in front of a stove, which is the main focus of the video. The kitchen is well-lit and clean, with white cabinets and a marble countertop. The girl appears to be excited and ready to start cooking. The style of the video is casual and family-friendly, with a focus on the girl's enthusiasm for cooking. The video is likely aimed at children and families, encouraging them to try cooking at home."
+    ],
+    "video_ids": [
+        "JAsKTr296Ik_12_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young girl in a pink chef's hat and apron, standing in front of a stove.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young girl wearing a pink chef's hat and apron, standing in front of an oven (which can be considered a stove in a kitchen context). The background includes kitchen cabinets and countertops, which aligns with the setting. There are no elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young girl wearing a pink chef's hat and apron, standing in front of a electric,which is be considered a stove in this kitchen context). She girl includes a elements and a, which are with with the setting of The is no elements in contradict the description description.\"\n}<|im_end|>\nGuidId",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..941fb17f62b8b14ebf9bf915399cff31d623aa42
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b440badae7abcc52fce894953e83c63b6fbfb771ce3b501ee2d29604db9127a
+size 147207
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..4c616f7cd831b687e5868eb929bc763fd11e4a2d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a baseball player in action, swinging a bat with intensity. The player, dressed in a blue jersey with the number 13 and the name \"BARRELL JR.\" on the back, is the main focus of the video. The bat, held firmly in the player's hands, is in motion, indicating a powerful swing. The player's stance and the position of the bat suggest a high level of skill and precision. The background is a blur of spectators, indicating that the scene is taking place in a crowded stadium. The overall style of the video is dynamic and energetic, capturing the excitement and tension of a live baseball game."
+    ],
+    "video_ids": [
+        "tV5mvaYOPLI_3_0to135"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Baseball player, bat? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a baseball player in a blue uniform with the name 'Sorriell Jr.' and number '13' on the back, holding a bat in a batting stance. The player and bat are the central focus, and there are no elements that contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a baseball player in a blue uniform with the number \\\"B.ianoelli'.' and the '13' on the back. holding a bat. a batting stance. The player is the are the central focus, and there are no elements that contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8c04d10d05763dcb62003c9b9330aa237ad9b1cb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1129-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd6a67aaffc7d0804443b8b28f8b2fb62adb4e6d30c14031d3466a2396366caf
+size 233213
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..0d40714176f63adc57e113d91fc262777d21afec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a skull-shaped cake as the main subject, which is placed on a silver platter. The cake is intricately decorated with gold and black colors, giving it a festive and celebratory appearance. The skull is adorned with a knife, suggesting a playful or edgy theme.  The cake is surrounded by various candies and chocolates, adding to the festive atmosphere. The candies are scattered around the cake, creating a visually appealing contrast with the cake's dark colors.  The background of the video is a white brick wall, which provides a neutral backdrop that allows the cake and candies to stand out. The wall is decorated with geometric shapes, adding an artistic touch to the overall scene.  The video is shot in a close-up style, focusing on the cake and candies. The camera angle is slightly elevated, providing a clear view of the cake and its surroundings. The lighting is bright and even, highlighting the details of the cake and candies.  Overall, the video captures a festive and celebratory moment, with the skull-shaped cake as the centerpiece. The combination of the cake, candies, and geometric shapes creates a visually appealing scene that is both playful and artistic."
+    ],
+    "video_ids": [
+        "LfWQdY11Zf8_2_0to153"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Skull-shaped cake, silver platter, candies, chocolates.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a skull-shaped cake decorated with gold patterns, placed on a silver platter surrounded by candies and chocolates. The core elements described are all present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a skull-shaped cake with with intricate and, placed on a silver platter. by various and chocolates. The elements elements of in present present and match depicted in the image.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7bf5e1b88d8759d9f5910285984a458aef0d237c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:35d03572eb4647ea47290c16477fa1c12be72bddb6a3e753d8348e9a3ce4befa
+size 96150
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..d41b1f85cf1c58a2ed986b27c9d78c002a5410e7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a young woman sitting in a room, engaged in a conversation with another person whose face is not visible in the frame. The woman is wearing a black and white patterned sweater and has her hair pulled back. She is looking directly at the other person, suggesting a serious or intense discussion. The room appears to be a living room or a similar indoor space, with a window and a bookshelf visible in the background. The lighting in the room is soft and natural, indicating that the video was likely taken during the day. The style of the video is realistic and naturalistic, with no visible special effects or artistic manipulations."
+    ],
+    "video_ids": [
+        "NDSxhgwiGQQ_40_0to181"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman and another person (face not visible).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman facing another person whose face is not visible, matching the 'Object(s)' condition. The young woman is clearly visible, and the other person is partially visible from behind, with their face obscured, fulfilling the description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman sitting another person whose face is not visible. which the descriptionObject(s)' condition. The woman woman is wearing the, and the other person is partially visible from the, fulfilling their face obscured. which the requirement.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ad61e0a37fcee7c2b88e49d63f634c983019ac3b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:33bd45570d6d1ebb34c1a7fad8c9bc9f53aa49a20b4f9484f424c483561c5b4a
+size 209619
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..19b7219f7cfe3fe2f8522e5ee20e578d65a06a84
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a black and white fifth-wheel camper being towed by a black pickup truck. The camper is large and has a black and white design with blue accents. The truck is a black pickup with a crew cab and a short bed. The truck is pulling the camper on a paved lot. The camper is being towed behind the truck. The truck is driving on the lot. The camper is being pulled by the truck. The truck is pulling the camper. The camper is being towed by the truck. The truck is pulling the camper. The camper is being towed by the truck. The truck is pulling the camper. The camper is being towed by the truck. The truck is pulling the camper. The camper is being towed by the truck. The truck is pulling the camper. The camper is being towed by the truck. The truck is pulling the camper. The camper is being towed by the truck. The truck is pulling the camper. The camper is being towed by the truck. The truck is pulling the camper. The camper is being towed by the truck. The truck is pulling the camper. The camper is being towed by the truck. The truck is pulling the camper. The camper is being towed by the truck. The truck is pulling the camper. The camper is being towed by"
+    ],
+    "video_ids": [
+        "Pn2cNHOSd1c_27_0to148"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black and white fifth-wheel camper, a black pickup truck with a crew cab and a short bed.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black and white fifth-wheel camper attached to a black pickup truck. The truck appears to have a crew cab and a short bed, matching the description. The overall scene is consistent with the specified objects without any clear contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black and white fifth-wheel camper being to a black pickup truck with The truck has to have a crew cab and a short bed, which the description. The scene scene is consistent with the provided objects, any additional contradictions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1f5ba7ec995ffd7624fdef1bdfcfc3b4bfd2db45
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:210bc0ca6b1af2942c5f1ef573cc9e96cafc6958fda28f66ec62c1d76b4bb45f
+size 163571
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..920b89ef0fda477d5de2245578db014ed31f8f47
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a black and red motorcycle wheel with a black tire and red rim. The wheel is positioned on a gravel road, and the motorcycle is stationary. The background features a rocky cliff with sparse vegetation, and the sky is clear and blue. The style of the video is a straightforward, static shot with no movement or action, focusing on the details of the motorcycle wheel."
+    ],
+    "video_ids": [
+        "cMHezRztzR4_12_0to150"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black and red motorcycle wheel with a black tire and red rim.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a scooter wheel, which has a black tire and a red rim, matching the description. The wheel is attached to a black frame, and the overall appearance aligns with the 'black and red motorcycle wheel' description, even though it's a scooter. The presence of additional elements like the gravel ground and background does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a motorcycle wheel that which is a black tire and a red rim, matching the description provided The wheel is part to a black frame, and the background design aligns with the givenblack and red motorcycle wheel' part. even though it is a scooter rather The presence of additional elements like the rocky and and the does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d19eec2477ebd75cfe7290b0bd31b126d3311f4e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76522549a6a12c4c07c16317031e38da914a892292568a73d1d230807588f89a
+size 254880
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..f4048027be6de7f209b42939c749fa4b3d2e4880
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen sitting at a table with a variety of hot sauce bottles in front of him. He is wearing a gray hoodie and appears to be in a dark room with a black background. The man is holding a bottle of hot sauce in his hand, and his facial expression suggests that he is about to taste it. The hot sauce bottles are of different sizes and colors, indicating a variety of flavors. The man's casual attire and the informal setting suggest that this could be a casual or informal event, possibly a hot sauce tasting or a challenge. The focus of the video is on the man and his interaction with the hot sauce, with the background serving to highlight the subject."
+    ],
+    "video_ids": [
+        "0-BQ_yTNyTs_128_163to296"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a table, and several hot sauce bottles.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man seated at a table with multiple hot sauce bottles arranged in front of him. The core elements described \u2014 a man, a table, and several hot sauce bottles \u2014 are all present and accurately represented. Additional elements (like a pitcher and logo) do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man sitting at a table with several bottles sauce bottles in in front of him. The man elements of in a man, a table, and several hot sauce bottles \u2014 are all present and accurately depicted in The elements,like the dark of a on do not contradict the core and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1dccc1144244c20a95a1b5379469c26548382b25
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:145d0f9d95a8d7a23cf274411b4104e0e35f6210efd82381f4736627d35124cd
+size 177069
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..ea9b68c581007eed1792440fadcd808885fd5293
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a black Ram truck on display at an auto show. The truck is parked on a white floor, and its shiny exterior reflects the bright lights of the showroom. The truck is the main focus of the video, with its sleek design and large wheels.  In the background, there are several people walking around, likely attendees of the auto show. They are dressed in casual attire, and some are carrying handbags. Their presence adds a sense of activity and liveliness to the scene.  The video is shot in a way that emphasizes the truck's design and features. The camera angles and lighting are carefully chosen to highlight the truck's details and make it stand out in the showroom. The overall style of the video is professional and polished, typical of auto show footage."
+    ],
+    "video_ids": [
+        "em2XqLk4lYU_9_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black Ram truck, several people? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a black Ram truck as the main subject, positioned prominently in the foreground. Several people are visible in the background, either standing or interacting near other vehicles, which matches the description. The presence of additional elements like other trucks or exhibition lighting does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a black Ram truck, the main object, which on on the center. Additionally people are visible in the background, standing walking or walking with the vehicles, which ful the description of The presence of additional elements like the vehicles and people stands does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6cfe67c166451ae3106e57246d1b8d57751236bc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5eb95e71b0c5003d5c37ed2baad61ae36ac1c127a9279e83bb3dd3b803eb1f72
+size 155237
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..7550dba9885996f151031220c499e96fa9a7e3f6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a kitchen, giving a thumbs-up gesture. She is wearing a black top and has blonde hair. The kitchen is equipped with a green door and a white wall. On the wall, there is a rack holding various kitchen utensils, including a pair of scissors and a spatula. The woman appears to be in a good mood, as indicated by her thumbs-up gesture. The overall style of the video is casual and friendly, with the woman appearing to be in a comfortable and familiar environment."
+    ],
+    "video_ids": [
+        "FHPhT2ldyEg_4_23to176"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a pair of scissors, a spatula? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman standing in a kitchen. A pair of scissors and a spatula are visible hanging on a rack in the background, fulfilling the 'Object(s)' condition. No conflicting elements are present that contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a woman giving in front room-like There pair of scissors and a spatula are visible on on the hook in the background, fulfilling the 'Object(s)' condition. The additional elements are present.\"\n would the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0b80407eadc6b7530b9913ecaa00c542cd50caa8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5656db4915d2fb7a27c1d953eebfcff1766a1972dfc22b630b2e2902cbe12eb7
+size 114107
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..81e13e44e2fc3c3645f8b11ea27b265f586220a3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit and tie, sitting in front of a blue background with the NBC logo. He is wearing glasses and appears to be speaking or listening intently. The style of the video is a standard news or interview segment, with a focus on the man and his expression. The background is simple and uncluttered, allowing the viewer to focus on the man and his message. The lighting is bright and even, highlighting the man's features and the NBC logo. The overall impression is of a professional and serious news or interview segment."
+    ],
+    "video_ids": [
+        "KP85i7bIpbs_8_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, wearing glasses, speaking or listening intently.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a suit, tie, and glasses, who appears to be speaking or listening intently. His attire and demeanor match the description, and the setting (CNBC interview) supports the context of him being engaged in a conversation. There are no elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed a suit and tie, and glasses, which appears to be speaking or listening intently. The posture and the align the description provided and there background suggestsaBC logo) supports the context of someone being in in a discussion or The are no additional in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bd6e6953fd6467eb11427c0ce7bd599c3c57cf3a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-113-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:295ed2111f3a18f66d856ff0dd1ac3b6a35e04642f67335859b509f9daaf1d2a
+size 163087
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..88bc787f22d20782f1f31927cabeeaf02f37cfe0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car, focusing on the back seat. The car has a sleek design with black leather seats. The back seat is empty, and the car appears to be parked. The car's interior is well-lit, and the seats are neatly arranged. The car's door is open, revealing the interior. The car's door is closed, and the interior is visible. The car's door is open, and the interior is visible. The car's door is closed, and the interior is visible. The car's door is open, and the interior is visible. The car's door is closed, and the interior is visible. The car's door is open, and the interior is visible. The car's door is closed, and the interior is visible. The car's door is open, and the interior is visible. The car's door is closed, and the interior is visible. The car's door is open, and the interior is visible. The car's door is closed, and the interior is visible. The car's door is open, and the interior is visible. The car's door is closed, and the interior is visible. The car's door is open, and the interior is visible. The car's door is closed, and the interior is visible. The car's door is open, and the interior is visible. The car's door is"
+    ],
+    "video_ids": [
+        "0L3ANDmGT5U_17_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Back seat, car's interior, car's door? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the back seat of a car, the car's interior, and the car's door. The focus is on the rear passenger area with the door open, revealing the seat and interior details, which aligns with the described conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows the interior seat of a car, the interior's interior, and the car's door. The image is on the back passenger area, the door open, which the seats and the details. which aligns with the ' '.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..31491dbb884c432a1380d2ae77c190d4b7950344
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:decb66ddaa0b0cbc01e90add846efc52a49442ed4f94acf214baf095aac91f72
+size 53354
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..8b4df6752bd97c339b854c36eb2e781827431ed5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a soccer match. The main focus is a soccer player, wearing a white and green jersey with the number 22, who is in the midst of an intense play. He is seen running across the field, his mouth open as if he's shouting or calling out to his teammates. His body language suggests he's fully engaged in the game, possibly strategizing or reacting to the ongoing play.  The background reveals a large stadium filled with spectators, their faces a blur of anticipation and excitement. The stadium's vastness and the crowd's energy add to the high-stakes atmosphere of the match. The lighting suggests it's either late afternoon or early evening, casting a warm glow over the scene.  The style of the video is a blend of action and sports photography, capturing the player in motion with a shallow depth of field that keeps him in sharp focus while the background blurs. This technique emphasizes the player's movement and the intensity of the moment. The video is likely shot from the sidelines, providing a wide view of the field and the crowd."
+    ],
+    "video_ids": [
+        "8BY9Z2omjJE_48_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A soccer player in a white and green jersey with number 22.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a white and green jersey with the number 22 clearly visible on the front. The jersey also features a crest and sponsor logos, consistent with a professional soccer uniform. The player's pose and the background of a stadium with spectators further confirm the context of a soccer match, fulfilling the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a white and green jersey with the number 22. visible. the back. The player also features green green on the logos, which with a professional soccer uniform. The player is attire and the setting, a soccer further spectators further support the context of a soccer match. align the ' description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ea32acbd7972f385a03a3195a66a2404f98ac18f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75df5214a05a5463e3da5430fced8d483c33751e3ba8c14b4d45b97fd1db81b7
+size 306314
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..0c4230a56c3ae079ec5962666d9e1aed4dfa28a7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man in a suit and tie walking across a basketball court. He is dressed in a dark suit and a striped tie, and he is looking down at the ground as he walks. The basketball court is filled with players, and the crowd in the stands is watching the game. The man is walking towards the camera, and the crowd is in the background. The video is shot in a realistic style, capturing the man's movements and the atmosphere of the basketball game."
+    ],
+    "video_ids": [
+        "ArhWQ--m-OQ_16_0to169"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a dark suit and striped tie, a basketball court, players, and a crowd in the stands.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a man in a dark suit and striped tie standing on a basketball court, with players and a crowd in the stands visible in the background. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a man in a dark suit and striped tie walking on a basketball court. with players and a crowd in the stands visible in the background. The elements elements of in present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f7310c91ce67d8a8e8cd30d4208a6d4d18790246
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8bf0d951d6949bcd146228522fb2f5aaa87d38641776367694be1aed2ce4ee03
+size 185542
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..dabec537569c6207136f2a30357a7c7f39e98ac7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with short brown hair, wearing a black and white striped dress, standing in a bustling city street. She is looking directly at the camera with a serious expression on her face. The background is filled with colorful lights and signs, indicating a lively urban environment. The woman is the main focus of the video, and her position in the frame suggests she is the subject of the video. The style of the video is realistic, capturing the everyday life of the city."
+    ],
+    "video_ids": [
+        "AGXyXmujyvw_12_78to213"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with short brown hair, wearing a black and white striped dress.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with short brown hair and a black and white striped dress, which matches the description. The background elements, such as blurred lights and pedestrians, do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with short brown hair wearing she black and white striped dress. which matches the description provided The background is, such as the lights and people, do not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..947ce6ad9c02d586495c88623e8a5e04e3ed1d7e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e2cd7a35c5b540421bc8312837cc45ec1acdcaa869798afa70ede2ae6692e0e
+size 317747
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..a477aa59af6f70aada899499ee2ffed95aefaf04
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a green military tank driving through a grassy field. The tank is large and imposing, with a yellow sign on its side that reads \"706 633 6072\". The tank is moving towards a sign that reads \"TANK TOWN USA.com\", indicating that it is heading towards a destination. The field is lush and green, providing a stark contrast to the tank's military green color. The sky above is overcast, suggesting that the weather is cloudy. The overall style of the video is realistic, capturing the tank's movement and the surrounding environment in detail."
+    ],
+    "video_ids": [
+        "oahvWAaydfk_16_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A green military tank, a yellow sign on the tank reading '706 633 6072', another sign reading 'TANK TOWN USA.com'.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a green military-style vehicle (likely an amphibious tank or armored vehicle) with the yellow number '706 633 6072' painted on its side. In the foreground, there is a yellow sign that reads 'TANK TOWN USA.com'. These elements match the description provided, and no contradictory elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a green military tank tank thattank a armoredious tank) a vehicle) with a number sign '706 633 6072' on on its side. There the foreground, there is a sign sign with reads 'TANK TOWN USA.com'. The elements match the description provided, indicating there conflicting elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f2ac818b0c72a37227482a7d742c7494c246919b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a6a94bfb8356d8919080db9be6ab0ffa13aed93f01118cefed1c333263364de
+size 269094
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..7a1487a2745995e1a5ec81752493225bf39b93b9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a bouquet of flowers in a pink vase placed on a white surface. The bouquet consists of various types of flowers, including pink, red, and yellow ones, as well as some greenery. The flowers are arranged in a way that they fill the vase and spill over the edges slightly. The vase is positioned in the center of the image, and the background is a plain white wall. The style of the video is simple and straightforward, focusing on the beauty of the flowers and the simplicity of the setting. The lighting is soft and even, highlighting the colors of the flowers without creating harsh shadows. The overall impression is one of tranquility and natural beauty."
+    ],
+    "video_ids": [
+        "1CegHAsM7bw_5_0to140"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bouquet of flowers in a pink vase.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a bouquet of colorful flowers (pink, red, and yellow) arranged in a pink vase, which matches the core description. Additional elements like a skull and candle are present but do not contradict the main object described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a bouquet of flowers flowers arrangedpink and yellow, and yellow) arranged in a pink vase. which matches the description description of The elements like the white and a are present but do not conflict the main object description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e6015180b5c0ae9ffa1e570db94bf46da628f62f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c286dd275951a6ba06c156f92e050694dd93fe7bf6d697fbb53eede8273e6db
+size 91048
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..01fe284d17aaaf3e0032592b71da80db7bf28790
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man standing next to a black and green motorcycle. The man is wearing a black jacket with red accents and a scarf around his neck. He is standing on a paved driveway with a building and a body of water in the background. The motorcycle is parked on the driveway, and the man appears to be posing for the camera. The style of the video is casual and seems to be a personal or promotional video for the motorcycle. The focus is on the man and the motorcycle, with the background serving as a scenic backdrop."
+    ],
+    "video_ids": [
+        "FabanK0pDho_20_44to180"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a black and green motorcycle, a black jacket with red accents, and a scarf.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a man standing next to a black and green motorcycle. The man is wearing a black jacket with red accents and a scarf, as described. All specified objects are present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as depicting showing a man wearing next to a black and green motorcycle. The man is wearing a black jacket with red accents and a scarf. which described. The the elements are present and match depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..90108ee12c42f93c07a2ba46ffe67b18ec38a59c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c7a5dba94ac6f9abcfa59265fc18cdc3718c65a3738455620a48232403b481ee
+size 106828
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..b3dc82a8a5415f2060c1695f263b3f93a20abf0c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a polar bear in an enclosure, partially submerged in water. The bear is seen interacting with its environment, moving its head and body slightly, and occasionally splashing water with its paws. The enclosure features a glass barrier through which the bear is visible, with water droplets forming on the glass, indicating recent movement or interaction with the water. In the background, there are rocks and some greenery, suggesting a naturalistic setting designed to mimic the bear's habitat. The lighting appears to be natural daylight, enhancing the visibility of the bear's white fur and the blue tones of the water. The bear's actions are calm and deliberate, showcasing its natural behavior in a controlled environment."
+    ],
+    "video_ids": [
+        "c1048204197074ce0b71ae4ced4a7687246a52b3e86df6b41348d44e221a016d"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A polar bear, water, rocks, greenery, glass barrier? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a polar bear in water, with rocks and greenery visible in the background, and a glass barrier separating the bear from the viewer. All elements mentioned in the condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a polar bear, a, with rocks and greenery in in the background. which there glass barrier is the bear from the viewer. The the mentioned in the ' are present and do depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dd544baaf80595546e9e8d8de3cf9406be2ebd0c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1130-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae269812099644207813935c7143fb7234eabeea3d401c8c8ddf9dbbf981269b
+size 235398
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..be72704cd2cba91ba505141f569f83b781c85af3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a red sports car in a spacious indoor setting. The car is sleek and modern, with a convertible top that is currently down. The car is positioned in the center of the frame, moving from left to right. The car's design is highlighted by its black rims and red accents. The background is a large, open space with a gray floor and a white wall. The lighting is bright and even, casting a soft glow on the car and the surrounding area. The overall style of the video is sleek and modern, with a focus on the car's design and the indoor setting."
+    ],
+    "video_ids": [
+        "HPfenX7t8ys_31_0to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red sports car with a convertible top down.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red sports car with its convertible top down, matching the description. The car is stationary in a studio-like setting, and no elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red sports car with a convertible top down, which the description provided The car is parked, a well-like environment, and the additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..71d954b8bccd04b023445bb7838281c7832fd3c2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:401aa9c59d436f06f4848bcff0ce1f7a52fbb717b1a2a6c1e3c093d1367059bd
+size 116447
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..f32cb66b3c35985aa48ba5f4ef572e2b52cc81b5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man standing in a kitchen, looking directly at the camera. He is wearing a gray sweater and has a beard. The kitchen is cluttered with various objects, including a refrigerator, a sink, and a potted plant. The man appears to be in the middle of a conversation or an activity, as he is looking at the camera with a slight smile on his face. The style of the video is casual and informal, with a focus on the man and his surroundings. The lighting in the kitchen is bright, and the colors are vibrant, creating a warm and inviting atmosphere."
+    ],
+    "video_ids": [
+        "Qmwey79SHN4_94_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a gray sweater with a beard, standing and looking directly at the camera.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a gray sweater and having a beard, standing and looking directly at the camera. These elements are consistent with the description, and additional background elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a gray sweater with has a beard, standing and looking directly at the camera. The elements match consistent with the description provided and there elements elements like not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ca013f3186e974afe43718ab7c8e0e07752e2e9b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba7036790a0de0ee3a99da9b4dd30e27fc8454d20dbd7e9da24cb87c61b89ee9
+size 175103
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..63451a6940b318a512ac6df43962fa643f9787fb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a delicious breakfast spread on a white plate, set against the backdrop of a wooden table. The plate is filled with a variety of breakfast items, including a sunny-side-up egg, crispy bacon, sausage links, and a side of white rice. The egg, with its perfectly cooked yolk, is placed next to the bacon, which is cooked to a golden brown. The sausage links, with their savory aroma, are arranged next to the egg. The white rice, with its fluffy texture, completes the meal. The wooden table provides a warm and rustic setting for the meal. The video is shot in a close-up style, focusing on the details of the food and the textures of the ingredients. The colors of the food contrast beautifully with the white plate and the wooden table, creating a visually appealing scene. The video does not contain any text or narration, allowing the viewer to focus solely on the food and the setting. The overall style of the video is simple and straightforward, with a focus on the food and the setting."
+    ],
+    "video_ids": [
+        "s5HcTZgEpmc_4_0to177"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Sunny-side-up egg, crispy bacon, sausage links, white rice? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a sunny-side-up egg, crispy bacon, sausage links, and white rice on a plate. These items are the main focus of the video and are accurately represented, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a plate-side-up egg, crispy bacon, sausage links, and white rice on a plate. The elements match the core focus and the image and match positioned represented as fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a4782bddb71728e4bf7084a8d8616bc885810976
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26af6d11ce16fec802fdb72746071acadcb029cae52b6059d5d9ca4967e9d7bb
+size 94395
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..1c4f889a2c527026192c3c48a234a7a53b1ff0d9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a white Mercedes-Benz SUV on display at an auto show. The car is positioned on a white platform, with a large screen in the background displaying a scenic road. The car is the main focus of the video, with its sleek design and shiny exterior reflecting the lights of the showroom. The camera captures the car from different angles, highlighting its features and design elements. The showroom itself is spacious and well-lit, with other cars on display in the background. The video is a promotional piece, aimed at showcasing the car's design and features to potential buyers."
+    ],
+    "video_ids": [
+        "FKO36vVFawM_4_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: White Mercedes-Benz SUV? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a white Mercedes-Benz SUV, which matches the specified object. The vehicle is clearly visible and occupies the central focus of the frames, with no contradictory elements that would negate this identification.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a white Mercedes-Benz SUV, which is the description object. The car is displayed displayed in is a central focus of the frame, fulfilling no conflicting elements present would suggest the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dde0008602211e193e6dc4166a0cc0362aa6f82b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8971072f275790797f83eb1737ea2bdaff86b7c60587f0f69cd31d6682980b47
+size 106788
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..0679537c88ef2cc16e7bf8a5010cdbb7d5fae606
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a large orange teddy bear sitting on a gray couch in a room. The teddy bear is wearing a silver necklace and has a black nose. In the background, there is a television mounted on the wall and a white bookshelf. The room appears to be a living room or a bedroom. The teddy bear is the main focus of the video, and it is shown in different positions on the couch. The style of the video is casual and simple, with no special effects or music. The teddy bear is the only moving object in the video, and it is shown in a relaxed and comfortable position on the couch."
+    ],
+    "video_ids": [
+        "4oziXx50Zfw_9_67to263"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large orange teddy bear? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a large orange teddy bear, which is the central object and matches the description exactly. The bear is shown being placed on a couch and then sitting there, fulfilling the 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a large orange teddy bear, which matches the central object in matches the description provided. The bear is sitting in held on a surface, is sitting upright, which the 'Object(s)' condition as any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c35b71e3faa624d2e2af5f2abb1e05fac5405dc0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f0613cb2621c80d55a641b454021fe82ef1c894bae3121cb4c127ffcae23995
+size 145616
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..84d2b7286dc650773391d4d1e98da0b9bc80cc37
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a single mallard duck gracefully swimming across a calm body of water. The duck, with its distinctive green head and white chest, is the focal point of the footage. It moves smoothly through the water, creating gentle ripples that spread outwards. The water's surface reflects the surrounding environment, including the sky and some scattered leaves, adding to the tranquil atmosphere. The camera follows the duck from behind, maintaining a steady focus on it as it swims away. The background features tall, dry grasses that frame the scene, enhancing the natural setting. The overall mood of the video is peaceful and undisturbed, highlighting the beauty of nature in a quiet moment."
+    ],
+    "video_ids": [
+        "5c0a546c6e531a6791e7c10027676134eae2037832325f82eb57b734073c7f4c"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A single mallard duck with a green head and white chest.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a single mallard duck swimming in water, which matches the description of having a green head and white chest. The duck's distinctive green head and white chest are clearly visible throughout the video, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a single mallard duck with in a. which matches the description of a a green head and a chest. The duck is features features head and white chest are clearly visible, the frames, and the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..67abc1146b37226011223ab46dedad67512b3b97
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:71c07336bd9e33e9b9284387c093ca47f4293e4751c1dfba72c8bc681a26eeb6
+size 163751
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..8decac3f716c860233a1424081da1e143c4dda7b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, an elderly woman with gray hair is seen sitting on a green leather couch in a room with a white wall. She is wearing a blue shirt and is holding a white mug with a blue handle. The woman is sipping from the mug, which has a purple flower design on it. The room has a wooden cabinet in the background. The video captures a simple, everyday moment of the woman enjoying her drink."
+    ],
+    "video_ids": [
+        "2GwLpaqgF5A_63_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An elderly woman, a green leather couch, a white wall, a wooden cabinet, a blue shirt, a white mug, a purple flower design? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an elderly woman with gray hair, wearing a blue shirt, holding a white mug with a purple flower design. She is seated on a green leather couch, with a white wall and a wooden cabinet visible in the background. All elements mentioned in the description are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts an elderly woman sitting gray hair, sitting a blue shirt, sitting and white mug with a purple flower design. She is sitting on a green leather couch in with a white wall and a wooden cabinet in in the background. The the in in the description are present and match with the video content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fc6b4a8c0524bdd2190331fb01ddc49c7b2c1adc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95652266c7c4cb9dc3529a04cd7a3b888b3063ba04a8d5664791bf3dc76bfdcd
+size 135875
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..aaa9e9b90885fb0ec86117e5d5d81378be0fc4a4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman with tattoos and a black turtleneck stands confidently in front of a body of water, with a tree and a rocky shore in the background. She is the central figure in the image, and her presence is commanding. To her left, a man with a beard and a floral shirt is seated, smiling at her. He appears relaxed and amused. To her right, another man with a mustache and a colorful jacket is seated, holding a book and looking at her with interest. The setting is serene and natural, with the water and the rocky shore providing a beautiful backdrop. The overall style of the video is casual and relaxed, with the three individuals appearing to be enjoying a leisurely day outdoors."
+    ],
+    "video_ids": [
+        "AE9UgQ74FSk_125_22to152"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with tattoos and a black turtleneck, a man with a beard and a floral shirt, another man with a mustache and a colorful jacket.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with visible tattoos and wearing a black turtleneck, seated between two men. One man has a beard and is wearing a floral shirt, while the other has a mustache and a colorful jacket. All three individuals match the description provided, and no elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a woman wearing a tattoos and wearing a black turtleneck dress a next two men. One man has a beard and is wearing a floral shirt, while the other man a mustache and is colorful jacket. The elements individuals are the description provided, indicating the additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b5f2e11b53830239db6c2e091b3ef7d446b86657
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1131-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6124ef6d4ae5629eae78e341e2ad1e9822f915a62de23976fc1262b9b6ccbe7
+size 226469
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..47dc04f0cb37075c5f6714d91e45c169c3b8bade
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a red sports car, showcasing its luxurious design and advanced technology. The car's interior is predominantly black, with red accents adding a touch of sportiness. The dashboard is equipped with a large touch screen display, providing access to various controls and features. The steering wheel is also black, with red stitching that matches the car's exterior. The car's seats are upholstered in black leather, providing a comfortable and luxurious seating experience. The car's door panels are also black, with red accents and a red door handle that matches the car's exterior. The car's interior is well-lit, with ambient lighting that highlights the car's design and features. The video is shot from the perspective of the passenger seat, providing a comprehensive view of the car's interior. The car's interior is immaculately clean and well-maintained, reflecting the car's high-quality craftsmanship and attention to detail."
+    ],
+    "video_ids": [
+        "MB_3KUNY9tE_42_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dashboard with a large touch screen display, steering wheel, seats upholstered in black leather, door panels, and a red door handle.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dashboard with a large touch screen display, a steering wheel, and seats upholstered in black leather with red stitching. The door panels are visible, and the door handle is red, matching the car's exterior color. All elements described in the condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a car with a large touch screen display, a steering wheel, seats seats upholstered in black leather. red stitching. The door panels are visible, and the red handle appears red, matching the description's interior color. The elements described in the condition are present in match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ea45dee8929579275c299579318cedae4f7e4104
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f81086a114eafbc7d38e5cc1d69b4cff8b5f2e6985eea77501d7f0da1b37af41
+size 176114
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..c7fdea310e41467a870a9a05d8933c75245bf118
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a bald man holding a camera with a large lens. He is wearing a green shirt and appears to be in a room with a desk and various items on it. The man is looking directly at the camera with a serious expression. The style of the video is a close-up shot of the man, focusing on his face and the camera he is holding. The lighting in the room is dim, with the man being the main source of light. The overall mood of the video is serious and focused."
+    ],
+    "video_ids": [
+        "1BZfgG5d7gk_13_0to176"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bald man, a camera with a large lens, a green shirt, a desk, various items on the desk? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man holding a camera with a large lens, wearing a green shirt, and seated at a desk with various items visible on it. The background also includes framed pictures and other objects, which aligns with the description without contradicting any core elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man holding a camera with a large lens, wearing a green shirt. and standing at a desk with various items on in it. The elements includes includes a pictures, a desk, which ares with the description of contradicting it of elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..47e41d1b8b88e4ee5138dde147559f5f2bd894aa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:573d84d6b31c22603b42dc8ca54c95b75f822e809f331ad85ce5ebc6b9c80f77
+size 71078
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..48fc845fbdfcb3661f27e0b9da772b46c836abd9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red Kia car driving down a city street. The car is in motion, with its headlights on, indicating that it is either dusk or dawn. The car's license plate reads \"P 309 OP 799\". The street is lined with parked cars and trees, and there are other vehicles visible in the background. The overall style of the video is realistic, capturing the everyday scene of city life."
+    ],
+    "video_ids": [
+        "Qd1qfhkI884_44_49to176"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red Kia car, parked cars, trees, and other vehicles in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red Kia car moving forward, with parked cars, trees, and other vehicles visible in the background. The scene matches the described elements without contradiction, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a red Kia car, on on which parked cars and trees, and other vehicles visible in the background. The presence is the description elements, any.\"\n fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d852d84b909dc31526544367abab80efdcb19758
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:16a35ac1855c22fa908f8c233514a2bf4ee9026a7b5105b9bb1166028ff1ab2d
+size 180812
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..37a0433634b05f49b2263b849be670adc138271e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in a forested area. He is wearing a green tank top and has a beard. The man appears to be looking off to the side, possibly at something or someone out of frame. The forest around him is lush and green, with trees and foliage filling the background. The lighting suggests it might be daytime. The man's expression is serious, and he seems to be deep in thought or focused on something. The overall style of the video is naturalistic, capturing the man in his environment without any obvious staging or artificial elements."
+    ],
+    "video_ids": [
+        "Hlg6XQUEvAQ_64_0to188"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man wearing a green tank top with a beard, looking off to the side.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The man in the video is wearing a green tank top and has a beard, and he is looking off to the side. These elements match the description provided, even though he is also holding a leaf and there is foliage in the background, which does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video in the video is wearing a green tank top and has a beard. which he is looking off to the side. The elements match the description provided. making though the is not wearing a phone, the are a in the background, which is not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a6532f7c1d597cbe906fea27e71003208c4c196c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:78625625c083da76b4336b6fa776abaf187c954c447a9c33f932212122485986
+size 141309
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..3ff1319008d5e47faa67ce0f6feb1f12cf234fd0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car, focusing on the steering wheel and dashboard. The steering wheel is black with a silver logo in the center. The dashboard features a touch screen display, which is turned on and displaying a menu. The car's interior is beige, and the steering wheel is positioned on the left side of the frame. The car appears to be a modern model, with a sleek design and a focus on technology. The video is likely a promotional or informational video, showcasing the car's interior and features."
+    ],
+    "video_ids": [
+        "B46vXkSXJOc_16_53to211"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, dashboard with a touch screen display? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a steering wheel with the Toyota logo and a dashboard featuring a touch screen display. These elements are central to the video and are accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a steering wheel with the Mercedes logo, a dashboard with a touch screen display. The elements directly consistent to the description and align consistent represented, any.}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e7e287ad3a4aa37639b9e216ae0367edfaf64339
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:726b65edc7951d30ad03ab8a28e2b41d72ef6ffffb6357c92cafd8621110486a
+size 131630
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..db172205b56a53f83333b88ed7de2e39ca331d63
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a luxury car, focusing on the driver's side. The car's interior is sleek and modern, with a black color scheme. The steering wheel is in the center of the frame, with the dashboard and center console visible. The car's door is open, revealing the door handle and window controls. The car's interior is well-lit, with the sunroof open, allowing natural light to fill the cabin. The car's design and features suggest it is a high-end model, likely a sports car or luxury sedan. The video captures the car's interior in a way that highlights its design and features, providing a glimpse into the experience of driving this luxury vehicle."
+    ],
+    "video_ids": [
+        "6vq1kH65dNI_53_0to157"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, dashboard, center console, door handle, window controls, sunroof.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the steering wheel, dashboard, center console, door handle, and window controls. The sunroof is not visible, but since the video is taken from an interior perspective with the roof open, it is implied that the car has a convertible top rather than a fixed sunroof. The absence of a visible sunroof does not contradict the core description, as the video still fulfills the 'Object(s)' condition for the other listed items.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows the interior wheel, dashboard, center console, door handle, window window controls. The presenceroof is also directly in but the the question is focused from the interior perspective, the sun partially, it can reasonable that the sun has a sun top or than a sun sunroof. The presence of a sun sunroof does not contradict the description description of as the video is fulfills the 'Object(s)' condition by the other elements elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d3ebbf48d5b301b33d828567f2a58ca74746ffee
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:912a4506f429835dc9a3f27861c9409d9025a219dd3316d0c236c03ea52565f6
+size 159497
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..c1ace5d2cff9f5547ad7aa60d8b8e4e31d44f2d4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a large tortoise in a grassy area. The tortoise is seen in three different positions: first, it is walking on the grass, then it is seen lying down, and finally, it is seen walking again. The tortoise is a large, slow-moving creature with a thick shell. The grassy area is lush and green, providing a natural habitat for the tortoise. The video captures the tortoise's movements and the peaceful environment in which it lives."
+    ],
+    "video_ids": [
+        "HG67p0ozd0M_21_65to208"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large tortoise? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a large tortoise (specifically, a snapping turtle) as the central subject, resting on green grass. While there are people and trees in the background, they do not contradict the core description of a large tortoise being the main focus. The tortoise is clearly visible and occupies the majority of the frame, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a large tortoise inlikelyally, a tort turtle) in the main subject. which on a grass. The the are no in a in the background, they are not detr the core description of the large tortoise. the main focus. The presenceoise's clearly visible and occupies a majority of the frames in fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b81e734ec9e1d1ae13cc7f8077c64100d06790c1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7dc30edb5c6bf628587f1f066776ac5a0393e821c5f20accc0dff96c883549f0
+size 259031
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..9a5727624562d1589d45f6792af79f43b3847398
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of joy and celebration, featuring a man and a woman standing together in a crowd. The man, dressed in a suit and tie, is waving to the crowd with a warm smile on his face. The woman, wearing a yellow jacket, is standing next to him, also smiling and waving. They are surrounded by a group of people, all of whom appear to be in a good mood. The crowd is diverse, with people of various ages and backgrounds. The setting appears to be an outdoor event, possibly a political rally or a public gathering. The atmosphere is lively and energetic, with the crowd cheering and applauding. The man and woman seem to be the main focus of the event, drawing the attention of the crowd. The video is a snapshot of a moment of unity and celebration, capturing the spirit of the event and the enthusiasm of the crowd."
+    ],
+    "video_ids": [
+        "OMA7-h8kCAE_2_94to229"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman standing together in a crowd.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a woman standing together in a crowd, which matches the 'Object(s)' condition. The man is Barack Obama, and the woman is Michelle Obama, both prominently featured in the foreground. The background contains other people, forming a crowd, which aligns with the description. There are no elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a woman standing together in a crowd. which align the 'Object(s)' condition. The man is wearing Obama, and the woman is a Obama, as of featured in the crowd. The crowd is a people, indicating a crowd, which iss with the description.\"\n The are no elements in contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..acfe22b76637522f240cda115475b7ba64380062
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1132-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:db8c507c11e26a97c9045bd66ae60b06f4f7b8f6cc43b8ca89453ecf1a7cb7ff
+size 227995
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..3a9b5510f3d55d5c22f76350788150f8f8c46b9e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video begins with a close-up view of a hand holding a clear glass tube filled with colorful liquids. The tube is part of an intricate setup on a concrete surface, which includes several other transparent containers and objects. As the camera pans to the right, it reveals more of the scene, showing a small glass jar with a cork stopper containing what appears to be a preserved insect or specimen. Adjacent to the jar is a tall, rectangular container made of clear acrylic, segmented into different colored sections, creating a vibrant display of hues. Further along, there is another transparent container holding multiple colorful straws or tubes, adding to the visual complexity.  The background features a parked car and a wall, suggesting an outdoor setting, possibly near a workshop or a creative space. The lighting is natural, casting shadows and highlighting the reflective surfaces of the objects. The camera movement is smooth, providing a detailed exploration of the arrangement without any abrupt transitions or shifts in focus. The overall composition suggests a meticulous and artistic"
+    ],
+    "video_ids": [
+        "732ca33505d6559e7cb66234d031f583556c11cceba330824f682a73d087c2b0"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A clear glass tube with colorful liquids, several other transparent containers, a small glass jar with a cork stopper containing a preserved insect, a tall rectangular container segmented into different colored sections, and another transparent container holding multiple colorful straws or tubes.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a clear glass tube with colorful liquids, several other transparent containers, a small glass jar with a cork stopper containing a preserved insect, a tall rectangular container segmented into different colored sections, and another transparent container holding multiple colorful straws or tubes. All these objects are clearly visible and match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a clear glass tube with colorful liquids, several other transparent containers, a small glass jar with a cork stopper containing a preserved insect, a tall rectangular container segmented into different colored sections, and another transparent container holding multiple colorful straws or tubes. The these elements are present visible and match the description provided.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..97ac08fc04181c9b1b964b4598e11ae9856671bf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6ac94cfb76b294ee6aca5a700761195f6b7b7484b09979f4f13528ca35875e3d
+size 232680
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..4a416352d396db390466ca973a634bf56082251a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man standing in front of a road construction sign in a snowy environment. The man is wearing a black jacket with red accents and black pants. He has a beard and is looking directly at the camera. The sign is orange and black with a diagonal arrow pointing to the right. There is also a green sign with white text that reads \"DIAGNATED SIDEWALK\" and a white sign with black text that reads \"SIDEWALK CLOSED.\" The background features a chain-link fence and a yellow construction vehicle. The style of the video is a straightforward, real-life depiction of a construction site in a snowy environment."
+    ],
+    "video_ids": [
+        "3odt1wn8klI_9_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man, road construction sign, green sign, white sign, black jacket with red accents, black pants, beard? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man standing next to construction signs in a snowy setting. He is wearing a black jacket with red accents, black pants, and has a beard. There are visible road construction signs, including a green 'Designated Walkway' sign and a white 'Sidewalk Closed' sign. All elements described in the condition are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man standing in to a signs. a snowy environment. The is wearing a black jacket with red accents, black pants, and has a beard. The are road road construction signs, including a green signHALated Roadway' sign and a white signNoewalk'' sign. The the in in the question are present in match with the video content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..afcce9baece755d3de40f92a3b6e6ce662dd7e18
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e727074ffaf34353d9ebd77d517b40ec693bc2a601a82a6b45ba077c889a15b
+size 80226
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..0fe9ed49961e06774569ce539aaf5effc5f8982d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man with dreadlocks and a headband is seen in a conversation with a woman. The man is wearing a colorful, patterned shirt and a necklace, while the woman is dressed in a blue top. They are seated on a bench, with the man leaning towards the woman, indicating an intimate or engaging conversation. The background features a building with a dome and a street with power lines, suggesting an urban setting. The overall style of the video is casual and candid, capturing a moment of interaction between the two individuals."
+    ],
+    "video_ids": [
+        "4o9cfUqSg-A_204_0to138"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with dreadlocks and a headband, a woman in a blue top, a colorful, patterned shirt man, and a necklace.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with dreadlocks and a headband, wearing a colorful, patterned shirt and a necklace, and a woman in a blue top. These elements match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with dreadlocks wearing a headband, a a colorful, patterned shirt, a necklace. alongside two woman in a blue top. The elements match the description provided, the questionObject(s)' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..896ff28990b09b63dcd644efe4f558682dcd8deb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da76a0725bb87a07a613c7bea6ff136252be7a2448b9673a8025c45cb9b0ad63
+size 255851
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..d21169ca2037082df324a78635558d81915dc5c6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of an orange tabby cat with a fluffy coat and large, expressive eyes. The cat is positioned in front of a mirror, which reflects its image back at it. Throughout the sequence, the cat appears to be observing itself in the mirror, occasionally shifting its head slightly from side to side. Its ears are perked up, indicating curiosity or attentiveness. The background is a plain, light-colored wall, providing a neutral backdrop that keeps the focus on the cat and its reflection. The lighting is soft and even, highlighting the cat's fur texture and the subtle details of its facial features. There are no other significant objects or characters in the frame, ensuring that the viewer's attention remains solely on the cat's interaction with its reflection. The overall atmosphere of the video is calm and introspective, capturing a quiet moment of self-examination."
+    ],
+    "video_ids": [
+        "370e45c8e013d0e05a7293c4f63fe05a7754317d183b5480846e44beb3f997d3"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An orange tabby cat with a fluffy coat and large, expressive eyes.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an orange tabby cat with a fluffy coat and large, expressive eyes, as described. The cat is seen looking at its reflection in a mirror, and its features match the description throughout the frames.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a orange tabby cat with a fluffy coat and large, expressive eyes. which described. The cat's depicted in to its reflection in a mirror, which the fur, the description of the frames.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b7437be9d191c26b3bbb57e3ef3ba1dcab938e55
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0515dc3a27d0241198204945a5f66177d679fa885d68cfa48e233eacd3f60bac
+size 175558
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..0f52947e8a23a30ac83223250eff51281564c99c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a glass bowl filled with coffee beans. The coffee beans are green and appear to be freshly roasted. The bowl is placed on a wooden surface, which adds a warm tone to the image. The camera angle is slightly elevated, providing a clear view of the coffee beans and the bowl. The focus is on the coffee beans, with the background blurred, emphasizing the texture and color of the beans. The style of the video is simple and straightforward, focusing on the coffee beans without any additional elements or distractions."
+    ],
+    "video_ids": [
+        "jwxZNriJO_4_39_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Glass bowl, coffee beans? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a glass bowl containing coffee beans, which matches the described 'Object(s)' condition. The coffee beans are visible falling into the bowl and filling it, and the bowl's transparent glass material is evident throughout the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a glass bowl filled coffee beans. which matches the description objectsObject(s)' condition. The bowl beans are visible and into the bowl, filling it, indicating the glass is transparency nature is is also. the video.}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0b581c65dd0be1f3169340aa826da283a4e54fc8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2337d1122470229ea36e2eac5a0e6645a256261d26e05676bd2f92d0c07200cc
+size 110530
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..9757e9e99a942018c32bf954a27c6b4d018fd906
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a car's engine compartment, focusing on the Mercedes-Benz logo and the engine cover. The style of the video is a straightforward, unembellished documentation of the car's engine, likely for educational or promotional purposes. The engine cover is black with the Mercedes-Benz logo prominently displayed in the center. The video captures the details of the engine cover, including the textured surface and the logo's metallic finish. The engine cover is the main subject of the video, with no other significant objects or actions taking place. The video does not contain any text or additional graphics. The style of the video is simple and direct, focusing solely on the engine cover and its details."
+    ],
+    "video_ids": [
+        "mTESV-xzV8U_7_0to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Mercedes-Benz logo, engine cover? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the Mercedes-Benz logo prominently displayed on the engine cover, along with the 'AMG' branding, which aligns with the described objects. The engine cover is the central focus of the video, and no conflicting elements are present that contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a Mercedes-Benz logo prominently displayed on the engine cover. which with the textMG' branding, which iss with the ' '. The focus cover is also main focus of the video, and the additional elements are present.\"\n would the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9073b69bd88d7370815c25af01ebbe682f7e5f92
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1929b3ffd1112675c0725d71952b221c944b591a7d2397e32001ec5a2e5a9470
+size 135763
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..259110189e9dd3276b9581d95408203257333c77
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the majestic beauty of a mountain range, with each frame showcasing a different perspective of the same scene. The first frame provides a distant view of the mountains, their peaks shrouded in a blanket of clouds. The second frame offers a closer look, revealing the rugged terrain and the patches of snow that cling to the mountain slopes. The third frame brings the viewer even closer, highlighting the verdant greenery that carpets the lower elevations of the mountains. The video is a testament to the grandeur of nature, captured from a bird's eye view."
+    ],
+    "video_ids": [
+        "0taffwFmPU0_107_48to184"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Mountain peaks, clouds, rugged terrain, snow, verdant greenery? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts mountain peaks with rugged terrain, patches of snow on higher elevations, and verdant greenery covering the lower slopes. Clouds are also prominently visible, drifting around the peaks and adding to the natural landscape. All elements described in the condition are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful mountain peaks, varying terrain, clouds of snow, the elevations, and verdant greenery in the lower slopes. Thes are also visible featured, adding across the mountain, over to the scenic landscape. The these described in the ' are present and contribute with the video content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4636e66681f26c4a9fc569cd0568be37f9bf455d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec0a4bb77078c24fd40be8ad3498877d7a4998594edc7544e32e695993086952
+size 147437
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..736ffd0e2187fd968a4137c2e3e2c0a8b93ee50e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up of a plate of rice with vegetables, including onions, tomatoes, and peppers. The rice is cooked and seasoned, and the vegetables are cut into small pieces. The plate is white, and the food is presented in a way that makes it look appetizing. The style of the video is simple and straightforward, focusing on the food without any additional elements or distractions. The lighting is bright, highlighting the colors of the vegetables and the texture of the rice. The video does not contain any text or additional graphics. The overall impression is that of a simple, yet delicious meal."
+    ],
+    "video_ids": [
+        "Rx6lRsV_UJc_2_0to146"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Rice, onions, tomatoes, peppers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a dish of rice mixed with visible pieces of tomatoes, green and yellow peppers, and sliced onions. These elements are consistent with the 'Object(s)' condition described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a dish that chopped, with chopped pieces of onions and onions peppers red peppers, and onions onions. The elements match consistent with the descriptionObject(s)' condition provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6259c04678ba1796a9eb92b818faf8c5e3fa87c2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1133-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:378df4c47c07bb2758c26db3b6d83e878673af9ebf141dd9f3a6d54e9979751e
+size 56244
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..3288a74ae55dcaf7c0d0510b651f46bf438b6faa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features an elderly woman named Denise, who is the main subject. She is seen in three different frames, each capturing a different moment of her life. In the first frame, she is seen with a surprised expression on her face, as if she has just heard some unexpected news. In the second frame, she is seen with a thoughtful expression, perhaps reflecting on the news she just heard. In the third frame, she is seen with a smile on her face, indicating that she is happy or content with the news. The video is a snapshot of Denise's life, capturing her emotions and reactions to different events."
+    ],
+    "video_ids": [
+        "rohIx95yzNw_15_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Elderly woman named Denise? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features an elderly woman with white hair, visible wrinkles, and a name overlay that clearly identifies her as 'Denise'. Her appearance and the text label match the specified condition of an elderly woman named Denise.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a elderly woman with styled hair styled wearing in, and a floral tag that reads identifies her as DeniseDenise'. The attire and the context description match the description ' of the elderly woman named Denise.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..94be14122e14ddb84b17183b4fa4de3f515ff464
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26ba5a836d1541f81ca2a2fcbcd1f780cf7b12ebd1ceeaa3ce799d0591b0003c
+size 143425
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..2b46a4730816ed87807e0dbdf4da5190d6e97809
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene moment in a European city, where two individuals are ascending a set of stairs. The stairs, adorned with a black railing, are nestled between two white buildings, each boasting balconies and potted plants. The buildings, with their white facades, stand out against the backdrop of a clear blue sky. The individuals, dressed in casual attire, are seen walking up the stairs, their figures framed by the buildings on either side. The overall atmosphere of the video is calm and peaceful, capturing a typical day in a European city."
+    ],
+    "video_ids": [
+        "U3b06ILXMus_28_0to131"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two individuals, stairs, black railing, two white buildings, balconies, potted plants? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing two individuals ascending a set of stairs with a black railing. The stairs are flanked by two white buildings, each featuring balconies with decorative railings. Potted plants are visible along the stairway and on the buildings, adding to the scene's authenticity. No elements contradict the description, and additional details like the bell tower and sky do not interfere with the core objects listed.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing two individuals walking a set of stairs. black black railing. The scene are flanked by two white buildings, each with balconies and p pings and Potted plants are visible on the balconway and on the balcon' adding to the scene's detail. The elements contradict the description, and the elements like the clear-shaped and the do not detr with the core elements mentioned.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d71f9007499587c7f6605ba0e04acbe1982d60d2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:484274280d14d39f21b90e3055b3819bce808aa179bb37a2a0ef6a3f850f726e
+size 262142
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..474b1fe7a4f0ba5032a06ab13006604f7e7e7895
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a black suit and glasses, sitting at a desk with a microphone in front of him. He appears to be speaking or giving a speech. The setting is a formal room with wooden walls and a wooden desk. The man is the main subject of the video, and he is the only person visible. The style of the video is a straightforward, professional recording, likely for a news or political event. The focus is on the man and his speech, with no additional elements or distractions. The lighting is even, and the camera is positioned to capture the man's face and upper body clearly. The microphone is placed close to the man, indicating that his speech is the primary focus of the video."
+    ],
+    "video_ids": [
+        "QHj7AQWS6eI_1_111to291"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black suit and glasses, sitting at a desk with a microphone in front of him.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black suit and glasses, seated at a desk with a microphone positioned in front of him. These elements align with the core description provided. Additional text overlays and logos do not contradict the primary visual elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black suit and glasses, seated at a desk with a microphone in in front of him. The elements match with the description description provided, The elementsural or background are not contradict the main description elements and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6e5706d0a84b1ff0e5299c237c6f3a29fb3cce15
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a52f09328d43b0e9d234a7178624a02a0f2c673c9da907d835f4eb776dcd3b4
+size 101880
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..c57afcdc6d24b54d6a618137e18e175cce2c1f30
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in front of a blue display case filled with various cleaning supplies. He is holding a large green bottle of cleaning solution, which is prominently displayed in the center of the image. The man is dressed in a white shirt with orange stripes, and he appears to be presenting or demonstrating the product. The background is a simple blue wall with white text that reads \"CleanItSupply.com\", indicating that the video is likely an advertisement or promotional material for the website. The overall style of the video is straightforward and informative, with a focus on the product being showcased."
+    ],
+    "video_ids": [
+        "CPkyE3gj2r0_0_0to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a blue display case, various cleaning supplies, a large green bottle of cleaning solution, a white shirt with orange stripes.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man in a white shirt with orange stripes, standing in front of a blue display case filled with various cleaning supplies. He is holding a large green bottle of cleaning solution. All elements mentioned in the description are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man wearing a white shirt with orange stripes, standing in front of a blue display case filled with various cleaning supplies. The is holding a large green bottle of cleaning solution. The elements in in the description are present, accurately depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..16307ee4f154567ccdb7a22e88b3df6e9f405005
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:051d299b37acf07cd242e491dd0d64457227dc3f889eba908b40fc5250810429
+size 123300
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..53658265193cdc70ed31ae712bfa18e71856dce2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a cityscape from a high vantage point, showcasing a mix of modern and traditional architecture. The city is nestled near a body of water, with boats visible in the distance. The buildings vary in height and design, with some featuring red roofs and others boasting glass facades. The streets are lined with trees, adding a touch of nature to the urban environment. The overall style of the video is aerial, providing a bird's eye view of the city and its surroundings. The colors are vibrant, with the blue of the water contrasting against the greenery of the trees and the various hues of the buildings. The video is a dynamic representation of urban life, capturing the essence of the city's architecture and its proximity to the water."
+    ],
+    "video_ids": [
+        "xkR0WEvd_5A_44_213to384"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Buildings (various heights and designs, some with red roofs, others with glass facades), trees, boats, water.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing various buildings of different heights and designs, including some with red roofs and others with glass facades. Trees are visible throughout the urban landscape, and boats can be seen docked near the waterfront. The water body is prominently featured in the background, completing the described elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing buildings buildings of different heights and designs, some some with red roofs and others with glass facades. There are also throughout the scene landscape, and there can be seen oned or the water. The presence body is also featured in the background, completing the scene elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..56f9f3c82f462a12cf12c0be1f2d9ef713a3a547
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:73e06f276925f49760b3391d4697af18c8ee5a9a842b993d74acc2a96840f6a5
+size 289679
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..5a79ae5dfdbc5fffa7ee6ebb9e575029c39a0cd4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a vibrant garden scene, showcasing a variety of flowers in full bloom. The garden is lush with green foliage, providing a rich backdrop for the colorful flowers. The flowers, in hues of orange and yellow, are arranged in a neat row, creating a visually pleasing pattern. The video is shot from a low angle, emphasizing the height of the flowers and the depth of the garden. The overall style of the video is naturalistic, capturing the beauty of the garden in its entirety."
+    ],
+    "video_ids": [
+        "tlaPndLlOVs_7_0to189"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Flowers in hues of orange and yellow? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features flowers in shades of orange and yellow, which are clearly visible throughout the frames. These flowers are surrounded by green foliage, but the core description of orange and yellow flowers is accurately fulfilled.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video predominantly features rows in hues of orange and yellow, which align the visible in the frames. The flowers are the by green foliage, and the focus focus of the and yellow flowers is met represented.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9257f821f3ba4840496bbb569a0acd8ed25c4903
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8efc79f250568cc581acce76335c9b41e87f0ca9a53e3e5d4526ad16215e47ba
+size 165816
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..ab648369c7a616721acf773a8cb81dfd220f146c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of celebration between two soccer players on a field. The player in the foreground, with a broad smile on his face, is embracing the player in the background, who is bald and wearing a jersey with the name \"ZIDANE\" and the number \"5\" on it. The players are wearing white jerseys, and the field appears to be wet, suggesting that it has been raining. The background is blurred, but it seems to be a stadium filled with spectators. The overall style of the video is dynamic and captures the excitement of the moment."
+    ],
+    "video_ids": [
+        "_pnkN6PO22g_11_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two soccer players? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two soccer players in white jerseys celebrating on a field. One player has long hair and is smiling broadly, while the other, with a bald head, has his back turned and is raising his arm. Both are wearing jerseys with visible names and numbers, consistent with soccer players. The scene is set in a rainy environment, which is typical for outdoor sports, and there are no elements contradicting the description of two soccer players.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two soccer players, a jerseys,, a field, The player is the hair and is smiling,, while the other is with a bald head, is the back to to is wearing his arm in The are wearing soccer with the numbers and numbers, which with the players. The setting is set in a stadium stadium, which is a for soccer sports events and the are blurred elements thating the description of two soccer players.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..065cd9ed4a50baeb523d68e7ca574057daf33772
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb5fe0f81412af43d913769f04fb57b65bd68e6f62b31e86d5f22dfd264e9438
+size 185165
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..486ca958616c94204c0daf7bbe927d79ea140775
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young man with a unique style, set against a brick wall backdrop. He has long, wavy hair and a pierced septum. His attire consists of a black jacket and a black and white scarf. The lighting in the video is predominantly purple, creating a moody and atmospheric effect. The man's expression is serious, and he appears to be deep in thought. The overall style of the video is artistic and introspective, with a focus on the man's appearance and the dramatic lighting."
+    ],
+    "video_ids": [
+        "70F3jt5lX8c_0_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man with long, wavy hair and a pierced septum wearing a black jacket and a black and white scarf? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man with long, wavy hair, a pierced septum, wearing a black jacket and a black and white scarf. These elements match the description provided, even though the lighting and background may be stylized, they do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a young man with long, wavy hair and which sept septum, wearing a black jacket, a black and white scarf. The elements match the description provided, making though the video and background are not stylized or which do not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a37f9010190ae7c99177cc44252634f6c66994d0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1134-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6cb538dd1dceb1af0b484581cc1ea5e64101f7b388ee6de0f6a7fb360614a54
+size 122753
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..4489e30a6601ff33417d43c7b861ecbbffcae9a7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a compilation of three frames featuring a man in a suit standing on a soccer field. The man appears to be a coach or manager, as he is dressed in a formal suit and tie. The first frame shows him standing alone on the field, looking off into the distance. The second frame shows him with a group of soccer players, who are all wearing blue jerseys. The third frame shows him again standing alone on the field, but this time he is gesturing with his hands, possibly giving instructions or making a point. The style of the video is a montage, with each frame capturing a different moment in time. The man's attire and the presence of the soccer players suggest that the video is related to a soccer match or practice. The overall tone of the video is serious and focused, reflecting the intensity of the sport."
+    ],
+    "video_ids": [
+        "7PE4-yiIxmQ_31_0to195"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, a group of soccer players wearing blue jerseys? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man in a suit and tie prominently in the foreground, and in the background, there is a group of soccer players wearing blue jerseys, which matches the description. The presence of other elements, such as text overlays and a soccer match scene, does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit and tie standing standing the foreground, and a the background, there is a group of soccer players wearing blue jerseys. which align the description provided The presence of the elements like such as the on and a soccer ball setting, does not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8074670608abf754514aa395ce3f5ca02461c5f9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7509d29df3f3c84b717bc6894b03ca06d3d551641deaf1395a53e9a9cbe1e6ca
+size 206192
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..31580f7661fa5010ddc2e2c84bfe3523ce7d7cf5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a culinary journey, featuring a close-up of a dish being prepared and served. The dish is a square of tofu, topped with a rich brown sauce and garnished with a sprig of green. The tofu is placed on a white plate, which is set on a wooden table. The tofu is the main focus of the video, with the camera zooming in on it as it is being prepared and served. The sauce is poured over the tofu, and the green garnish is added, creating a visually appealing dish. The wooden table provides a rustic backdrop to the dish, adding to the overall aesthetic of the video. The video is shot in a realistic style, capturing the details of the dish and the preparation process with precision."
+    ],
+    "video_ids": [
+        "Y0anchqXI0A_16_295to423"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Square of tofu, rich brown sauce, sprig of green garnish, white plate, wooden table? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a square of tofu topped with rich brown sauce and a sprig of green garnish, all presented on a white plate that sits on a wooden table. The core elements described are clearly visible and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a square of tofu, with a brown sauce and garn sprig of green garnish, all placed on a white plate. is on a wooden table. The sequence elements of in present present and match represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d50af9163d6c36f83edc75f320ade73dc478ad18
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4f53df440a2aece48ff98a567a420659037c09ee88a80bbb9129b5dcff3a28eb
+size 108901
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..7b36ecddc7e60971b9f89dab66f89699c136f9fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man named Rep. Jim Welch, a Democrat from Springfield, who is speaking in an office setting. The office is decorated with framed pictures on the wall, and the man is wearing a blue shirt and tie. The video is a news segment, as indicated by the red and blue banner at the bottom of the screen. The man appears to be addressing an issue or making a statement, but the specific content of his speech is not visible in the image. The overall style of the video is professional and informative, typical of a news broadcast."
+    ],
+    "video_ids": [
+        "-i9rjxeYSmI_9_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man (Rep. Jim Welch), a blue shirt, a tie, a red and blue banner at the bottom of the screen.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man who is identified as Rep. Jim Welch via the on-screen banner. He is wearing a light blue shirt and a tie. A red and blue banner is visible at the bottom of the screen displaying his name and affiliation. These elements match the core description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man wearing appears wearing as Rep. Jim Welch, the text-screen text. He is wearing a blue blue shirt and a dark, The red and blue banner is present at the bottom of the screen, text name and other. The elements match the description description provided.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..69ee9b06fb18759f11a377038e5a221544099a01
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d460cf60a9f8db9b3868094ae018034876a454071cb51ec3474fc7f01acbcc8
+size 131869
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..9f7e82bdfa54ebb2e60f9826ecbcfddca6256510
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment of a football player in action. The player, wearing a blue and white uniform with the number 12, is seen in three distinct frames. In the first frame, the player is seen in a running stance, his body leaning forward in anticipation. The second frame captures the player in mid-air, his arms outstretched as he prepares to make a catch. The third frame shows the player landing, his body crouched in a defensive stance. The player's helmet, with its red and white color scheme, stands out against the backdrop of the football field. The video is a freeze-frame of a high-energy moment, capturing the intensity and excitement of the sport."
+    ],
+    "video_ids": [
+        "cavGkO4wR9E_0_0to196"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Football player, helmet? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a football player wearing a helmet, matching the 'Object(s)' condition. The player is in full uniform, including the helmet with a faceguard, and is visibly celebrating. There are no elements that contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts football football player in a helmet and which the 'Helmet(s)' condition. The player is depicted a uniform, including a helmet, a face mask, which is engaged engaged or The are no contradictions in contradict the presence.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..77d1f4962f6bcae4b260d0626b87230e411f332a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2624061d258aefbdcb5652a83fc3400de65695ae7da87b3f64d609df34d658a4
+size 300898
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..c38f3f7a394d0c612c23a9509e0463fa648ada14
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a small brown puppy in a grassy field. The puppy is seen walking through the tall grass, its fur slightly matted from the outdoor elements. The puppy's eyes are wide and curious, taking in the surroundings. The background is a blur of green and red, suggesting a mix of trees and flowers. The puppy's ears are perked up, indicating alertness. The overall style of the video is a close-up, focusing on the puppy's face and upper body. The camera angle is slightly above the puppy, giving a sense of the puppy's height in relation to the grass. The video is a snapshot of a puppy's exploration of the outdoors."
+    ],
+    "video_ids": [
+        "9F7CAXiltLo_3_0to139"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small brown puppy? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a small brown puppy as the main subject, which matches the 'Object(s)' condition. Additional animated elements (like the red stick figure and cartoon dog) do not contradict the core description of the puppy being present and central to the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a small brown puppy in the main subject. which align the descriptionObject(s)' condition. The elements grass likelike the grass flowers and) theish) are not contradict the core description of the puppy. the.\"\n are to the video.\"\n}<|im_end|>\nGuidId",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c20eb612337ebf2800c16116c1e90951c6b43fa0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:db01a614ac2409f9376361469cb32ebd210025426a421cc5140a242c48105565
+size 220051
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..3fad879d5f7be7203d2383481e6bc96ad520bc5a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a luxury car, focusing on the steering wheel and dashboard. The steering wheel is prominently displayed, featuring the distinctive logo of the car manufacturer. The dashboard is sleek and modern, with various controls and displays. The car's interior is well-lit, highlighting the high-quality materials and craftsmanship. The video is shot in a close-up style, focusing on the details of the steering wheel and dashboard. The overall impression is one of luxury and sophistication."
+    ],
+    "video_ids": [
+        "jS_2BaTQhs8_28_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, dashboard? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a close-up of the steering wheel with the Lamborghini logo, which satisfies the 'Steering wheel' condition. The dashboard is partially visible in the background, fulfilling the 'dashboard' condition. No elements contradict these core descriptions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a steering-up view a steering wheel and the logoorghini logo in which is the 'Steering wheel' condition. Additionally dashboard is also visible in the background, showing the 'dashboard' condition. The additional contradict the descriptions descriptions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c22e58b5c5610db4d17b177afe9598814945b4c5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0cd7ed984acc64eec2629df6d3ed294c7615d4f4048c822a146bd8021b8a0700
+size 225249
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..b4b0e133083620a80de6346424c3e50f32556aac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a sports talk show featuring two men. The host is seated at a desk with a laptop, a microphone, and a coffee mug. He is wearing a black jacket and glasses. The guest is seated in a black leather chair, wearing a blue jacket and a black beanie. He is holding a microphone and appears to be engaged in a conversation with the host. The background of the show is a basketball court with a scoreboard and a logo. The style of the video is a professional sports talk show with a focus on basketball."
+    ],
+    "video_ids": [
+        "9MvTKiJd9PA_14_19to187"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men - one is the host seated at a desk with a laptop, microphone, and coffee mug; the other is the guest seated in a black leather chair holding a microphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men in an outdoor setting. One man is seated at a desk with a laptop, microphone, and coffee mug, consistent with the role of a host. The other man is seated in a black leather chair, holding a microphone, fitting the description of a guest. The scene matches the core description without contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men in a interview setting, One man is seated at a desk with a laptop, a, and coffee mug, fulfilling with the description of a host. The other man is seated in a black leather chair, holding a microphone, which the description of a guest. The presence appears the core description provided any.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c31242bc7372a8640476f834beaeb0e9a1ed8524
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1677a44c3c2e917f3862769f6abc933f3e8dee4efaa648d077aa9628ab1891a8
+size 154724
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..af5182bc017f2bb4e60a273550b078a1dfdbe116
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a green shirt preparing food in a small kitchen. He is standing at a counter with various items on it, including a bowl, a cup, and a fork. The kitchen is compact and well-organized, with a window that lets in natural light. The man is smiling and appears to be enjoying the process of cooking. The style of the video is casual and homey, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "Iex31girZHs_2_51to264"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a green shirt, a bowl, a cup, a fork? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a green shirt, a white bowl, a pink cup, and a fork. These objects are present and consistent with the description provided. The presence of other elements, such as a dog and a TV, does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man in a green shirt, and bowl bowl, and yellow cup, and a fork. The elements match present and match with the description.. The man of additional kitchen like such as the microwave and additional window, does not contradict the core description and}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..efb42c1eec309ffa1ab532bcf4d568372a405a6b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1135-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e67151b85b381d79ec99c0ccba56196962490b10ed905f14de7f76cdecab716d
+size 145271
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..954994e282e63b5da2507c2eba33950b91b3fe91
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a heartwarming scene in a hospital room. A woman, dressed in a blue uniform, is seen tending to a young child who is lying in a hospital bed. The child, who is bald, is being comforted by the woman. The room is filled with medical equipment, including an IV drip and a heart monitor, indicating the child's need for constant care. The room is also adorned with pictures and posters, adding a personal touch to the sterile environment. The video beautifully captures the tender moments between the caregiver and the child, highlighting the importance of human connection in healthcare settings."
+    ],
+    "video_ids": [
+        "HqoTFWRniJo_203_23to159"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a blue uniform, a young bald child in a hospital bed, and medical equipment such as an IV drip and a heart monitor.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman in a blue uniform (likely a nurse), a young bald child in a hospital bed, and medical equipment including an IV drip and a heart monitor. These elements are present and accurately depicted, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a woman in a blue uniform interactinglikely a nurse) a young bald child in a hospital bed, and medical equipment such what IV drip and a heart monitor. The elements align consistent and match depicted, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a0a2fe168c0df31a8b98d887a862dcb2f445a6dc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:947a5b312e8f4d396b59afee524d657a854eff85f13589b0f353b7bffdbecd22
+size 178712
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..35550e52a17a11a2568f4d908f20338f5bda7c22
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen creating a colorful tree decoration using small beads. The person starts by holding a small red house-shaped object, which serves as the base of the tree. They then proceed to add a large cluster of multicolored beads to the top of the house, forming the tree's canopy. The beads are vibrant and varied in color, creating a visually appealing effect. The entire process takes place on a wooden table, which provides a natural and rustic backdrop for the activity. The person's hands are visible throughout the video, demonstrating the careful and precise movements required to create the tree decoration. The overall style of the video is simple and straightforward, focusing on the craft activity without any additional distractions."
+    ],
+    "video_ids": [
+        "VWGJN7qpHsQ_64_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small red house-shaped object, multicolored beads, and a person's hands.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a small red house-shaped object suspended by strings from a cluster of multicolored beads, which resembles a hot air balloon. A person's hands are also visible, interacting with the object. Additional elements like a basket and containers of beads are present but do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a small red house-shaped object with in mult, a mult of multicolored beads. which are a tree air balloon. The person's hands are visible visible, manipulating with the object, The elements like the wooden with a are beads are present, do not contradict the core description.}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b68d5f0e095a5a86434d14bf86ac73243ba71fcd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75247a82b747ab102e89fdd8011b595a23614ba8499f64a84eafdcf337cf4672
+size 188837
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..7f86259e248f2afc8f9f6ff69d3627a0d9974d6d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man with gray hair is seen in a pink shirt, sitting at a table with a woman in a white shirt. He is speaking into a microphone, suggesting that he might be giving a speech or presentation. The setting appears to be outdoors, with palm trees visible in the background, indicating a warm climate. The man's expression is serious, and he seems to be concentrating on his speech. The woman is listening attentively, adding to the overall atmosphere of the scene. The video captures a moment of public speaking, set against a tropical backdrop."
+    ],
+    "video_ids": [
+        "KDpL_A0aPwA_68_0to161"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with gray hair in a pink shirt and a woman in a white shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man with gray hair wearing a pink shirt, and a woman in a white shirt is visible in the background. These elements match the description provided, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows depicts a man with gray hair in a pink shirt and and a woman in a white shirt. also next the background. The elements match the description provided, indicating there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0034b2288ffd8b373c0d8d7a8a6da079e394576c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2989a10f0c87e2802321569f3d011b1a1f43e6ca46083dd8533626bd47e05eaf
+size 196386
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..18b35324ba45fb8122ca2b62cf7f8a437dc6e1ae
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a green cutting board with a knife, a whole purple onion, and a bunch of green parsley. The knife is positioned on the cutting board, ready to be used. The onion and parsley are placed next to each other, suggesting that they are about to be chopped. The cutting board is placed on a surface, and the background is blurred, focusing the viewer's attention on the cutting board and its contents. The style of the video is simple and straightforward, with a focus on the food preparation process."
+    ],
+    "video_ids": [
+        "za4JiukhSLI_1_0to150"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Knife, whole purple onion, bunch of green parsley? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a knife, a whole purple onion, and a bunch of green parsley on a green cutting board, matching the specified objects. The presence of a celery stalk does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a knife, a whole purple onion, and a bunch of green parsley on a green cutting board. which the description objects. There presence of these green stalk in not contradict the description description as}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cc6a8152ac828ed43648a1e871438fdcf82bc407
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:240346a91ba44ae447887a279d2f06337142dd2f03a190bde763ebdd9b1420f3
+size 149373
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..6a148976a1ea0a4c660e9ea9c6c36a9b48fb7a11
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a bustling city street scene. In the first frame, a man in a white coat is seen walking across the street, while a motorcyclist is seen riding past him. In the second frame, the man in the white coat has reached the other side of the street, and the motorcyclist is now further down the road. In the third frame, the man in the white coat is now walking down the sidewalk, and the motorcyclist has turned a corner and is no longer visible. The street is lined with tall buildings, and there are traffic lights at various points along the road. The overall style of the video is a realistic depiction of urban life, with a focus on the movement and interaction of people and vehicles in a busy city environment."
+    ],
+    "video_ids": [
+        "KSA-Sh8y46s_14_24to153"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a white coat, a motorcyclist? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a white coat standing at the crosswalk, and a motorcyclist (on a scooter) is visible crossing the street. These elements match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a man in a white coat and on a sidewalk and and a motorcyclist iswith a blue) is also in the street. The elements match the ' of,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3d58e93da66bb9da2947904c881945157eb09ac7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86b34ce6fdfe12f834c7bb67160c4278be65ff6f0e9d8982896a4d883ab33957
+size 292966
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..66420d34332852c143ab4efa61a0cb4922281acd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman with curly blonde hair is seen in a bakery. She is wearing a beige sweater and has a warm smile on her face. The bakery is filled with various pastries and bread, displayed on shelves in the background. The lighting in the bakery is soft and inviting, creating a cozy atmosphere. The woman seems to be enjoying her time in the bakery, possibly contemplating which pastry to choose. The overall style of the video is casual and relaxed, capturing a simple yet charming moment in the woman's day."
+    ],
+    "video_ids": [
+        "Hj7OFElSIlM_11_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with curly blonde hair wearing a beige sweater, a warm smile on her face.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with curly blonde hair wearing a beige sweater, and she has a warm smile on her face. The background elements (e.g., shelves with products) do not contradict the core description and are acceptable as additional context.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a woman with curly blonde hair wearing a beige sweater, and she has a warm smile on her face. The background appears,sh.g., shelves with items) do not contradict the core description and are acceptable as additional elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ec546a263a4fbb9e77554586257f75baed2347c9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c6d5c993c8ba1cecef951e68f349c581388c30b69981c4d6d1ca9cb1000fb0b
+size 328269
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..2b052d16ebd874a14d6adf6b3e5f5b8d55641acc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man in a suit and tie walking across a basketball court. He is dressed in a dark suit and a striped tie, and he is looking down at the ground as he walks. The basketball court is filled with players, and the crowd in the stands is watching the game. The man is walking towards the camera, and the crowd is in the background. The video is shot in a realistic style, capturing the man's movements and the atmosphere of the basketball game."
+    ],
+    "video_ids": [
+        "ArhWQ--m-OQ_16_0to169"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a dark suit and striped tie, a basketball court, players, and a crowd in the stands.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a man in a dark suit and striped tie standing on a basketball court, with players and a crowd in the stands visible in the background. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a man in a dark suit and striped tie walking on a basketball court. with a and a crowd in the stands visible in the background. The elements elements of in present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1098b92ee5b89b6b7167154513332cf77d16a2cf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1704f79427d6819aa43b1a34de6b70834c4a89057818f1e32838b1c640ba4f7
+size 250721
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..18a9331def2de5fed1d4671988304bcf6a36cd5f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a beautifully wrapped gift on a pink floral plate. The gift is tied with a brown twine bow and is adorned with a floral pattern. The plate is placed on a pink floral tablecloth, which adds to the overall aesthetic of the scene. The gift appears to be a box of soap, suggesting a sense of luxury and care. The video captures the essence of a thoughtful gift-giving moment, with the focus on the presentation and the attention to detail in the wrapping. The style of the video is simple yet elegant, emphasizing the beauty of the gift and the care put into its presentation."
+    ],
+    "video_ids": [
+        "wocypAxjRSo_41_21to157"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A beautifully wrapped gift, a brown twine bow, a box of soap.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a beautifully wrapped gift with a brown twine bow, which appears to be a box of soap wrapped in floral fabric. The visual elements align with the description, even though the exact nature of the contents (soap) is not explicitly confirmed, it is implied by the context and presentation.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a beautifully wrapped gift with a gold twine bow, which matches to be the golden of soap. in decorative-pattern and The presence elements in with the description, fulfilling though the bow material of the gift (soap) is not explicitly stated by the is reasonable by the context of the.\"\n}<|im_end|>\nGuidId",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5f0ab816816873256b8cc97e80d97bbfd98d8373
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1136-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dad325a9ab67f409f34457f8b9d9c37cc941f4b807a5ddbdbdd0635d67b1e550
+size 125242
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..c656367fe815e5de81aab7522cf4a364f2b22164
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a breathtaking scene of a mountainous landscape with a body of water. The water is filled with floating chunks of ice, indicating a cold climate. In the distance, a boat is seen navigating through the icy waters, adding a sense of adventure to the scene. The mountains rise majestically in the background, their peaks dusted with snow. The sky overhead is filled with clouds, casting a dramatic light over the scene. The overall style of the video is one of natural beauty and adventure, capturing the raw and untouched nature of the location."
+    ],
+    "video_ids": [
+        "a5xurq0pg8g_3_0to135"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A boat in the water, mountains in the background, and clouds in the sky.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a boat in the water, surrounded by floating ice, with prominent mountains in the background and clouds in the sky. These elements align with the described conditions without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a boat in the water, with by mountains ice, with mountains mountains in the background and a in the sky. The elements align with the ' ', any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4552858f88e6da4efcda4662a8be168c151d9930
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9712406f13d68a29a0d49a1e738a615d8b20bf0a1e7f194f7d0b8fb20c43983d
+size 212271
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..52f9d7ab2eebd9341a5cac92b6afeef58f7c1d18
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a delicious ham and cheese croissant on a brown plate with a floral pattern. The croissant is golden brown and flaky, with a filling of ham and cheese visible. The plate is placed on a pink table, and there are other objects in the background, including a pink toy and a white mug. The style of the video is casual and inviting, with a focus on the croissant as the main subject. The lighting is soft and warm, highlighting the textures and colors of the food and the plate. The background is blurred, drawing attention to the croissant and making it the focal point of the video."
+    ],
+    "video_ids": [
+        "F_CiNNoysM4_33_0to154"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Ham and cheese croissant, pink toy, white mug? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a ham and cheese croissant on a brown plate, along with a pink toy (a pink pig figurine) and a white mug with floral patterns in the background. These elements match the specified objects in the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a cro and cheese croissant on a plate plate with which with a pink toy anda small elephantgyine) and a white mug in a patterns. the background. The elements match the description objects in the description,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a4abe6900fd3050614d9bcad64b680c5ceb9c388
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57fbf4cebdf57191c45b7054e7fe3dfdcba7b9e800102253ad84721617751b23
+size 101149
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..aba328247579df5897d2c0d528909bb58effc8d3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the breathtaking view of a mountainous landscape, showcasing the grandeur of nature. The camera, positioned high above, provides a bird's eye view of the scene, allowing us to appreciate the vastness of the terrain. The mountains, with their rocky peaks and lush green vegetation, dominate the frame, their majestic presence undeniable. The perspective of the video, combined with the natural beauty of the landscape, creates a sense of awe and wonder. The image is a testament to the power and beauty of nature, captured in a single, stunning frame."
+    ],
+    "video_ids": [
+        "FOlnSV0QOe8_116_160to338"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Mountains, rocky peaks, lush green vegetation? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features towering mountains with distinct rocky peaks, covered in lush green vegetation. The visual elements align perfectly with the described conditions, showcasing a natural landscape dominated by these features without any conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features mountains mountains with rocky rocky peaks, which in lush green vegetation. The landscape elements in well with the ' ', showcasing a vibrant landscape that by these features. any conflicting elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f6cc8d8e80d2f25101841d0db9491e22386c9b82
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:352216b0be2b244d1c68c1a1a119543508fa724c1df0ac5dee5dd644e2ad9f86
+size 256681
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..abcca35e0479ac6732ed991ba33e3edb61fa5545
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse of a car's interior, showcasing the transition from day to night. The car is a modern, gray SUV with a sleek design. The interior is well-lit, featuring a clean and minimalist aesthetic. The dashboard is equipped with a touch screen display, and the steering wheel is adorned with various buttons and controls. The car's seats are upholstered in a light gray fabric, providing a comfortable and luxurious feel. The windows are rolled up, and the car is parked in a garage. The lighting in the garage gradually changes from bright to dim, indicating the passage of time. The car remains stationary throughout the video, and there are no people or other objects in the frame. The overall style of the video is calm and serene, with a focus on the car's interior and the changing light conditions."
+    ],
+    "video_ids": [
+        "0Lc9pnWOymQ_6_0to190"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Modern, gray SUV, touch screen display, steering wheel with buttons and controls, light gray fabric seats, closed windows? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows the interior of a modern SUV with gray upholstery, a visible touch screen display on the dashboard, a steering wheel with buttons and controls, and light gray fabric seats. All windows appear to be closed, and the overall scene matches the described elements without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts the interior of a car vehicle with a fabric, a touch touch screen display, the dashboard, a steering wheel with buttons and controls, and light gray fabric seats. The these appear to be closed. as the overall design is the description elements without any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..12d031fdb6479a205197d664eb75c1e30ae957af
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c31a919943def3b6542a24a30dc00fe75ad514fc14a30ca766685fa9e9bd700a
+size 156588
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..e187c5029409ec583d462a6dbbcd11c9ac7a7131
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man standing in front of a large blue sign with white text. He is holding a red microphone and appears to be speaking or announcing something. The sign behind him has the words \"Retro Mobile\" and \"F\u00e9vrier 2019\" written on it, suggesting that the event is related to vintage cars and is taking place in February 2019. The man is wearing a black jacket and a lanyard, indicating that he may be a staff member or an organizer of the event. The style of the video is straightforward and informative, likely intended to provide information about the event or to introduce the speaker."
+    ],
+    "video_ids": [
+        "8wJrPxbB2XE_0_0to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a large blue sign with white text, a red microphone? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man holding a red microphone, standing in front of a large blue sign with white text. These core elements are accurately represented in the video, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man holding a red microphone, standing in front of a large blue sign with white text. The elements elements match present represented in the video, fulfilling the 'Object(s)' condition as}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f84a4554d3d2f3885812e84c984a43656b3e2d85
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e11b00eeeba635c66f7289ab560accb39992ed382fcb2f75529e39fd7feafc0
+size 212024
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..87ec3fba6b4aa80b3d48dae0587a53d73a1e7f0c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up, high-resolution, and detailed view of an engine compartment. The engine is covered with a black plastic cover, and the hood is open, revealing the engine. The engine is shiny and well-maintained, with various hoses, belts, and other components visible. The engine compartment is clean and well-organized, with no visible signs of wear or damage. The style of the video is realistic and informative, focusing on the details of the engine and its components. The video is likely intended for educational or promotional purposes, showcasing the engine's design and features."
+    ],
+    "video_ids": [
+        "1QjGnf9jvN0_24_0to139"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Engine, hoses, belts, other components? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows various engine components such as hoses, belts, and other mechanical parts within the engine bay of a car. The detailed view includes fluid reservoirs, air intake tubes, and engine covers, all of which align with the described 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts an components components such as the, belts, and other mechanical parts. the engine bay. a car. The presence depiction of the reservoirs, engine intake systems, and other covers, which of which are with the description 'Object(s)' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..19c6dc9708281cccbe4c8ca57f1edb4aa21be0ac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb67eb916b27fc0412a5d02aaf0af0113337aec87ea0b0b5feb84426369945eb
+size 240599
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..9135730dfdc5b3626279d930a5cf10c15a3a207e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a car from the perspective of the driver's seat. The car's steering wheel is prominently featured in the foreground, with the key fob hanging from the ignition. The dashboard, featuring a radio and air conditioning vents, is visible in the background. The car's interior is black, and the key fob has a red tag with the number \"3\" on it. The style of the video is straightforward and unembellished, focusing solely on the car's interior without any additional context or action."
+    ],
+    "video_ids": [
+        "LFmGxz4kKD4_14_0to188"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, key fob, dashboard, radio, air conditioning vents? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the steering wheel with the Volkswagen logo, a key fob (red tag with the number 3), dashboard with gauges, a radio unit with a screen, and air conditioning vents. All specified objects are present and accurately depicted without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a interior wheel, a key logo, a key fob withred with with the number '3), the with variousuges, radio radio,, buttons display, and air conditioning vents. These these objects are present and match depicted, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..36d8742f27728574f436558e27d814db2a2459a5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1fd33207c0ddf55c5236067a068e63092f0f997a3bf48832d026956348f810be
+size 47850
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..b27253a6225707a7568ed219a5d4ebc5b6fca3c6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a conversation between two men at a car show. The man on the left, dressed in a dark suit, stands with his hands in his pockets, attentively listening to the man on the right. The man on the right, wearing a light-colored suit, is gesturing with his hands as he speaks, indicating an animated discussion. They are standing in front of a red SUV, which is the focal point of the scene. The car show is bustling with activity, with other people and cars visible in the background. The lighting is bright, illuminating the red SUV and the men, creating a lively atmosphere."
+    ],
+    "video_ids": [
+        "0_ajU5BeR38_3_0to155"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, one in a dark suit and one in a light-colored suit, and a red SUV? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men: one wearing a dark suit and the other in a light-colored suit, standing next to a red SUV. The scene is consistent with the described objects, and no elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two men, one in a dark suit and the other in a light-colored suit. standing next to a red SUV. The setting is set with the description elements, and there additional contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..00220981eb038cab40033386f5842d5aff32441b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1137-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5cb5cd7a6b609c1a913035bbf02b9ecd6b133893e7fe2347ecce59276fb0047a
+size 135148
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..2d9df41b4fc6b6d35bcbd2b290a1f152125b0216
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up scene of a large black beetle with prominent white markings on its head and thorax. The beetle is positioned on a ground surface covered with small pebbles and dirt. It is actively engaged in feeding on a smaller insect, which appears to be a grasshopper or a similar type of insect. The beetle uses its mandibles to grasp and tear at the prey, which is lying on its back. The beetle's movements are deliberate and focused as it continues to consume the insect. The background remains out of focus, emphasizing the beetle and its meal. The BBC logo is visible in the top left corner, indicating the source of the footage. The overall scene is a detailed depiction of a natural predatory behavior, showcasing the beetle's feeding process in a realistic and unembellished manner."
+    ],
+    "video_ids": [
+        "ccda7762ef131d163b0fc2b3ecac85d4864af6569e750ada559a315816c08bc9"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large black beetle with white markings, a smaller insect (likely a grasshopper).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a large black beetle with distinct white markings on its thorax and a smaller greenish insect, consistent with a grasshopper, on the ground. The beetle is actively interacting with the grasshopper, which aligns with the described objects. There are no conflicting elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a large black beetle with white white markings, its shellax and head smaller insect insect insect, which with the grasshopper, which the ground. The setting is interacting interacting with the smallerhopper, which aligns with the description scenario. The are no additional elements in contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5ea140074a9359122d8677a0e68757ecfdfc4cb2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:946ccb14219b47448756dbf5c114f90781c72170754e95d84619b57dd202836b
+size 186994
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..eb040ebd3c6bc2fc3626739ebee1522d30158f98
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a blue suit with a pink pocket square, standing in front of a brick wall with green plants on a shelf. He is gesturing with his hands and appears to be speaking. The style of the video is a close-up shot of the man, focusing on his facial expressions and body language. The background is simple and uncluttered, allowing the viewer to focus on the man. The lighting is bright, highlighting the man's features and the colors of his suit and pocket square. The overall impression is that of a professional or formal setting."
+    ],
+    "video_ids": [
+        "Gn_d63s0nWU_34_0to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue suit with a pink pocket square, gesturing with his hands.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue suit with a pink pocket square, and he is actively gesturing with his hands while speaking. These elements match the core description provided, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue suit with a pink pocket square, and he is gest gesturing with his hands. speaking. This elements match the description description provided, indicating there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0e0661278f1ae90150976a9a313ab0171d0ac422
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76ad81b305a67a26818bfccc163534f0d1acbebf0e9943e561f4ed87e8543808
+size 188524
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..447004e5016cb5cf46f1a8004a3519008a52712d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young couple is seen embracing each other in a warm and affectionate hug. They are standing in front of a brick building, which is adorned with palm trees. The man is wearing a blue shirt and a backpack, while the woman is dressed in a white shirt and a pink backpack. The couple's happiness is evident in their smiles and the way they hold each other. The video captures a moment of love and affection between the two, set against the backdrop of a beautiful day."
+    ],
+    "video_ids": [
+        "3dupaeI-7ZQ_14_132to290"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young couple embracing, a man in a blue shirt with a backpack, a woman in a white shirt with a pink backpack.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young couple embracing, with a man in a blue shirt and a backpack, and a woman in a white shirt with a pink backpack. Their actions and attire match the description, and additional elements (like another person walking by) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a young couple embracing, with the man wearing a blue shirt and a woman, and a woman in a white shirt with a pink backpack. The attire and attire match the description provided and there elements likelike the person in in in do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fdefaea0753c2cf30b0f47ac58010387f52d6d80
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fcfd8f026cd1edd99bb2cd4ade9c6e1b1d035182669363fdb41670b6c655e045
+size 182390
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..0da44336b74c0d4727e12cfeca67ed53e8882387
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a cooking tutorial featuring a chef in a kitchen. The chef is wearing a black chef's coat and glasses. He is holding a tray of pita bread. The kitchen is well-equipped with various cooking utensils and ingredients. The chef is preparing a dish that requires two pita breads. The style of the video is informative and instructional, with a focus on the chef's actions and the ingredients being used. The video is likely aimed at home cooks looking to learn new recipes."
+    ],
+    "video_ids": [
+        "84fr99urLW0_41_0to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chef in a black chef's coat and glasses, tray of pita bread, two pita breads for preparation.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chef wearing a black chef's coat and glasses, which matches the description. The chef is handling pita bread, and a text overlay confirms '2 pita breads' are being used for preparation. Although other kitchen items are visible, they do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chef in a black chef's coat and glasses, holding matches the description. The chef is holding aita bread, and there tray overlay suggests theChef pita breads for for being prepared for preparation. The the elements items are visible, they do not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3c25fafa1a2c267edcc29647fc13071255ac1d7b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c0ab77c68d5ad5618ced5c87f05f4ecd1683b883d45fe94d17672caf2d98d1b8
+size 130999
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..ecee9e4aa979463f426b3d70143d0442950b897a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman with blonde hair is seen in a kitchen setting. She is wearing a red apron and a gold necklace. She is engaged in a conversation with a man who is not visible in the image. The kitchen is well-lit and features a window with a view of the outdoors. There are several potted plants placed around the kitchen, adding a touch of greenery to the space. The woman appears to be in the middle of a discussion, possibly about a recipe or a cooking technique. The overall style of the video suggests a casual and friendly atmosphere, with the focus being on the woman and her interaction with the man."
+    ],
+    "video_ids": [
+        "8nNXWC8rIv8_22_0to103"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with blonde hair wearing a red apron and a gold necklace. A man (not visible).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair wearing a red apron and a gold necklace, which matches the description. The man is partially visible on the right side of the frame, though not fully in view, so the description 'A man (not visible)' is slightly misleading but not contradicted since he is partially visible. The core description is largely fulfilled.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair wearing a red apron and a gold necklace. which matches the description. The presence is not visible on the left side of the frame, but not fully in view, which he description ofA man (not visible)' is also misleading as still contradictoryicted. the is indeed visible. The presence elements is largely fulfilled.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..53ae1ca2e3a0535583db79cbc9705a78385b6ff8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:18db9e00d7e82af3c11da3f06ef54b053f15250e906d4c32e0a7f72e35d191e0
+size 159472
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..35a4466211ff8c90e21f900da2d830864827adc6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases the interior of a futuristic car, focusing on the driver's seat and dashboard. The car's interior is sleek and modern, with a minimalist design. The dashboard features a large, curved screen that displays various pieces of information, including the speed and time. The steering wheel is also curved and has a futuristic design. The driver's seat is white and appears to be made of a soft, comfortable material. The car's interior is well-lit, with a soft, ambient light that highlights the car's design and features. The video is likely a promotional video for the car, showcasing its advanced technology and design."
+    ],
+    "video_ids": [
+        "Yp7PmBRJaL0_34_0to187"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dashboard with a large, curved screen, steering wheel with a futuristic design, driver's seat made of white, soft material.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a modern car interior with a large, curved digital dashboard displaying various metrics, a futuristic steering wheel with a central logo and minimal physical controls, and a driver's seat upholstered in white, soft-looking material with orange stitching. All core elements described are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a car car interior with a large, curved dashboard dashboard screen information controls. a steering steering wheel with a sleek logo, aistic controls, and a driver's seat upholstered in white, soft material material. a accents. The these elements of in present and match depicted in}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5e848410d17f750e55aa560738dcf705b632fe37
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:33bf94384259732822fed3ad2dbfaaeec86c1c2d3ccd55e6e990b411b06c50cb
+size 182310
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..2a2065c3fdc1331c0688b91ffc9ba4afa00f6e9c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen enjoying a large portion of French fries in a red and white checkered paper container. He is seated in a car, with the window open, allowing a glimpse of the outside world. The man is dressed casually, and his focus is solely on the delicious fries he is about to enjoy. The video captures the simple pleasure of indulging in a favorite snack while on the go."
+    ],
+    "video_ids": [
+        "NRf6lZYzxU0_37_36to174"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a large portion of French fries, a red and white checkered paper container? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man sitting in a car, holding a red and white checkered paper container filled with a large portion of French fries. The core elements described are all present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man holding in a car, holding a large and white checkered paper container filled with a large portion of French fries. The man elements of in present present and accurately depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..12fed2aeed451cbc15d41970b73c9e893d1cf358
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f7d801705a0b7dae297f24a3d1af8666997cbed1373c847daf06c4378e74b65
+size 209173
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..3e0123bb1d481f0ace00a35c1975e3b023fd55db
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video presents an aerial view of a cityscape, focusing on a unique building that stands out due to its vertical garden design. The building is covered in lush greenery, with trees and plants growing on its facade, creating a vibrant contrast against the surrounding urban environment. The greenery appears to be well-maintained, with a variety of plant species contributing to the building's aesthetic appeal.  The building is situated in the heart of the city, surrounded by other structures that exhibit traditional architectural styles. The cityscape is dense, with buildings of various sizes and designs, indicating a bustling urban environment. The streets are lined with trees, adding a touch of nature to the urban setting.  The video captures the essence of the city's blend of nature and architecture, showcasing the innovative design of the vertical garden building amidst the more conventional cityscape. The aerial perspective provides a comprehensive view of the building's unique feature, allowing viewers to appreciate the harmony between urban development and environmental sustainability."
+    ],
+    "video_ids": [
+        "Xq1YjHL4a50_156_0to166"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A unique building with a vertical garden design, surrounded by other buildings and streets lined with trees.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a tall building covered in lush greenery, consistent with a vertical garden design. It is surrounded by other urban buildings and streets with trees, matching the description. The presence of additional elements like parked cars and pedestrians does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a unique building with in green greenery, which with a vertical garden design. The is situated by other buildings buildings, streets, trees lining which the description of The presence of additional elements like the cars and a does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0a3eff123c2ede39807dd97030d22fbe20dce8d3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1138-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50495bf08aecc9aab011bd71dfbea0c0a6dce76cc686291686ec28adb19b7c0f
+size 131763
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..5189e4fd32efde71b80a2334064c4307f8e6e27f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young man with a beard and short brown hair, wearing a gray suit and a white shirt with a purple pattern. He is seated in front of a blurred background that appears to be a green forest. The man is speaking, and his facial expression suggests he is engaged in a conversation or interview. The style of the video is a close-up shot, focusing on the man's face and upper body. The lighting is soft and even, highlighting the man's features without creating harsh shadows. The overall impression is that of a professional or formal setting, possibly a news interview or a corporate event."
+    ],
+    "video_ids": [
+        "S-8p97Ryk1s_12_22to161"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man with a beard and short brown hair, wearing a gray suit and a white shirt with a purple pattern.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man with short brown hair and a light beard, wearing a gray suit over a white shirt with a subtle purple pattern. The description matches the visual content accurately, with no significant contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man with a brown hair and a beard beard, wearing a gray suit and a white shirt with a purple purple pattern. The description matches the visual elements of, with no contradictions contradictions or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5052b5272921c0dc433d99b49dcaeb7d68d9fca7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6736c0080a8c346c1806f44fd1695ab9807e7a03afdddecc44278e1713e7ccfc
+size 248121
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..358af46df99dff42943891c9bf14c8977a3826cd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are standing in front of a small white airplane with red stripes. The man is wearing a white shirt with a blue collar and a face mask, while the woman is dressed in a red dress. The man is holding a camera and appears to be taking a picture of the woman. The airplane is parked on a tarmac, and the sky is clear and blue. The overall style of the video is casual and candid, capturing a moment between the two individuals."
+    ],
+    "video_ids": [
+        "5Zc38o_cd7A_141_39to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman interacting near a small airplane. The man is wearing a uniform and holding a device, while the woman is stepping down from the plane and smiling. Both individuals are prominently featured and the scene aligns with the 'Object(s)' condition of having a man and a woman.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man and a woman standing in a small airplane. The man is wearing a white that a a small to possibly the woman is dressed forward from the plane. wearing. The individuals are present featured, fulfill scene iss with the descriptionObject(s)' condition of the a man and a woman.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..105b6b95a641d1ebc4792d42891faa6207e2e2d0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e45682dddfd89bd0148045f6dab13f8b920f553ca8d22197dc58944d28ec00c5
+size 177016
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..96376d8caf6217e01383c39f7f74d290c2a9d8fd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up of a majestic lion's face, showcasing its powerful features and intense gaze. The lion's fur is a rich, golden color, and its eyes are a striking shade of amber. The lion's nose is a deep black, and its mouth is slightly open, revealing its sharp teeth. The lion's face is slightly dirty, with a noticeable wound on its forehead. The background of the video is a lush, green forest filled with tall trees and dense foliage. The lion's gaze is directed straight at the camera, creating a sense of connection between the viewer and the animal. The overall style of the video is realistic and naturalistic, capturing the lion in its natural habitat with stunning detail and clarity."
+    ],
+    "video_ids": [
+        "K0QFRV8m7AY_51_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A majestic lion's face? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a close-up of a lion's face, which matches the description of a 'majestic lion's face'. The lion appears injured with visible wounds, but this does not contradict the core description. The background foliage is present but does not interfere with the main subject.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a close-up of a lion's face, which align the description of a 'majestic lion's face'. The lion's to, visible wounds, which this does not contradict the core description of The focus is is present, does not conflict with the main subject,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..46652ecf46b7885e629d9714c65a89f38248ed12
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:058e544e91aeac773ab6133a0df72712b50e27a5e81b010f7e8e2cac020aed81
+size 258624
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..16403c112ca693cf0a8b8491cf338cec0de79c20
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with red hair standing in a classroom setting. She is wearing a green and blue patterned blouse and has her hands clasped together in front of her. The classroom is filled with rows of desks and chairs, all of which are empty. The woman appears to be speaking or presenting, as she is looking directly at the camera. The lighting in the room is bright, and the overall atmosphere is calm and quiet."
+    ],
+    "video_ids": [
+        "1E40kAiA1X8_3_388to514"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with red hair, wearing a green and blue patterned blouse, with her hands clasped together in front of her.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with red hair, wearing a green and blue patterned blouse, and her hands are clasped together in front of her. These elements match the description provided, and there are no contradictions with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with red hair wearing wearing a green and blue patterned blouse, and her hands are clasped together in front of her. The elements match the description provided, making there are no additional or the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b05b710bd508abd76a754b4349cecebef84d50f0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ad8ae5a818846eeae277f9dfff06dd75f0d759e4f89c15053fe80022e616b3cf
+size 110945
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..0985cba173f5018d8d9d7d020f2f2435b05628b5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are in a kitchen, preparing a meal together. The man is pouring something from a silver pot into a green bowl, while the woman watches attentively. The kitchen is well-equipped, with various appliances and utensils visible, including an oven, a microwave, and a sink. There are also several bottles and bowls on the counter. The scene is set against a backdrop of a wooden shelf filled with various items, including a potted plant and a vase. The overall style of the video is casual and homey, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "FsECdfbcoyg_18_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman standing in a kitchen setting, engaging in what appears to be a cooking demonstration. Both individuals are visible and central to the scene, fulfilling the 'Object(s): A man and a woman' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man and a woman in in a kitchen.. engaging in cooking appears to be a cooking activity. The individuals are present and are to the scene, fulfilling the 'Object(s)' A man and a woman' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..68152b76d0e1ed965d03e17efabbb62dd7f0c99d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c7b7fa81f097c5462e44bada06a65882a312df197c4be97808028dd86c52b4f5
+size 148287
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..9f91dddc366c218736c04f08136bc1c5c3bfe152
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a luxurious white boat with a sleek design, floating on a calm body of water. The boat features a spacious deck area with a comfortable white sofa and a matching white table. The table is equipped with a built-in ice bucket, suggesting that it's designed for relaxation and entertainment. The boat is docked near a pier, and the water reflects the boat's white exterior, creating a serene and inviting atmosphere. The overall style of the video is elegant and sophisticated, highlighting the boat's luxurious features and the tranquil setting."
+    ],
+    "video_ids": [
+        "hWQstWeGYaI_7_0to129"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A luxurious white boat, a spacious deck area, a comfortable white sofa, a matching white table, and a built-in ice bucket.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a luxurious white boat with a spacious deck area, a comfortable white sofa, a matching white table, and a built-in ice bucket (visible as a circular opening under the table). All elements described are present and consistent with the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a luxurious white boat with a spacious deck area. which comfortable white sofa, a matching white table, and a built-in ice bucket.part in a blue blue in the table). The elements described in present and match with the description depicted}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1f5e7c65c86afa1336d5c4143b8f0fca052076f9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cfe0b6e64c3dd7344cc042744d23901b465d2041d4cdfc95b8c18ee5ab8cc3d9
+size 200379
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..1590265ec036827885f9583cf19c1ed2904682b0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a tray of watermelon and honeydew melon slices on sticks, arranged in a visually appealing manner. The watermelon slices are cut into triangles and are placed in the center of the tray, while the honeydew melon slices are cut into rectangles and are placed around the edges of the tray. The tray is placed on a table, and the background is blurred, focusing the viewer's attention on the fruit. The style of the video is simple and straightforward, with a focus on the fruit and the presentation. The lighting is bright, highlighting the freshness and color of the fruit. The video does not contain any text or additional elements, and the focus is solely on the fruit and the tray."
+    ],
+    "video_ids": [
+        "_RwPm2WNqmQ_42_16to147"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Watermelon and honeydew melon slices on sticks.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows watermelon and honeydew melon slices on sticks, arranged on a tray. The watermelon slices are red with black seeds, and the honeydew melon slices are pale green. These elements match the description exactly, with no conflicting elements present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows slicesmelon and honeydew melon slices on sticks, which in a plate. The descriptionmelon slices are bright with black seeds, and the honeydew melon slices are green green. The elements match the description provided, and no additional additional present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b4761ce08db6c968635af59a46ed2a832a8f27d2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0e4fdc0c310afd03ec1a7badd755aa71d7dcbf85466a8169a92a67032ac55658
+size 57517
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..f4e2f49f96d3bf438b534e66d2c2b4ab95a8712d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a polar bear in its natural habitat, navigating through a rocky landscape. The bear, with its white fur, is seen walking on all fours, its body slightly hunched over as it moves. The bear's head is turned to the side, indicating that it is aware of its surroundings. The rocky terrain is scattered with small rocks and pebbles, adding texture to the scene. In the background, a body of water can be seen, suggesting that the bear is near a water source. The overall style of the video is realistic, capturing the bear in its natural environment without any human intervention. The focus is on the bear and its interaction with the environment, providing a glimpse into the life of this majestic creature."
+    ],
+    "video_ids": [
+        "K2ROd-hKw8Y_31_93to266"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A polar bear? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a polar bear standing on rocks near water, which matches the 'Object(s)' condition. The bear's appearance, size, and behavior are consistent with that of a polar bear, even though it appears wet and is interacting with colorful objects. These additional elements do not contradict the core description of a polar bear.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a polar bear walking on a near a, which align the descriptionObject(s)' condition of The bear's white, including, and the are consistent with the of a polar bear, and though the is to, slightly on with the rocks that The objects elements do not contradict the core description of the polar bear.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..922533e8e1150e75b9f0df3c12d8153748ef24ee
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1139-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1050123f2a258cae6659526acb784dd141ab7c57210d02f510aa47324881faf9
+size 356772
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..dc4c794b9b5a6b531a81aeb891fc7068cea45f79
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a character with a distinctive red and black helmet, which has a visor and a blue light on the front. The character is wearing a red and black outfit with a high collar and a chest piece that has a pattern of lines and curves. The background of the video shows a dark, industrial setting with metallic structures and pipes. The lighting in the scene is dim, with the character's helmet and outfit being the most illuminated elements. The style of the video is reminiscent of a science fiction or cyberpunk genre, with a focus on the character's costume and the industrial setting."
+    ],
+    "video_ids": [
+        "Yg8OM0Grlt0_13_0to140"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Character with a red and black helmet, blue light, red and black outfit with a high collar and chest piece with lines and curves.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a character wearing a red and black helmet with a blue light in the center, and a red and black outfit featuring a high collar and chest piece with distinct lines and curves. The visual details align closely with the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a character wearing a red and black helmet with blue blue light, the vis, which a red and black outfit with a high collar and a piece with lines lines and curves. The character elements match with with the description provided.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7ad2fa6ebfd4787d9acbad6af7328d21e5e3d112
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d7d48e67239488f9bec5f222a271a637f93f9712653ee8f7a06c2e979ef6b2dc
+size 222777
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..906a868bf8f55828b3389cdde637ed4f8dd7cc2e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red sports car parked on a concrete surface, with a gray car parked next to it. The red car has black rims and a sleek design, while the gray car has a more subdued appearance. The background features a building with a corrugated metal exterior and a few potted plants. The style of the video is straightforward, with no additional elements or actions taking place. The focus is solely on the two cars and their immediate surroundings."
+    ],
+    "video_ids": [
+        "rAk8REiy3DI_9_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red sports car, gray car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red sports car and a gray car parked next to each other, matching the 'Object(s)' condition. The camera pans out to reveal both vehicles, and no elements contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a red sports car and a gray car parked in to each other in which the descriptionObject(s)' condition. The presence angle across to reveal both cars, and the additional contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..980ccb0f487744414776227d17667a9088331d9b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a6a392f825d03b2f4681d5e46945e80d62808dc5b40c888a69ff1d3ea08f8356
+size 145604
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..b5caa359f205b059a768b58fbe91cc0a488a47cd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit and tie, standing in front of a large, ornate painting. He is gesturing with his hands, possibly explaining something or making a point. The painting behind him depicts a group of people, adding a sense of depth and context to the scene. The man's attire and the painting suggest a formal or professional setting. The overall style of the video is polished and professional, with a focus on the man's speech and the painting behind him."
+    ],
+    "video_ids": [
+        "XfywDp_spwo_20_21to239"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, a large, ornate painting.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man dressed in a suit and tie, holding a glass, and standing in front of a large, ornate painting with a gold frame. These elements match the core description provided. Additional elements, such as the patterned wallpaper and the 'byutv' logo, do not contradict the description and are acceptable as they do not conflict with the primary objects of interest.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a man in in a suit and tie, standing a microphone of which standing in front of a large, ornate painting. a group frame. The elements match the description description provided, The elements such such as the maned floor and the manH'iful' text, do not contradict the main and are acceptable as long do not conflict with the main elements mentioned a.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..57f362a15d9d26373186c6dd734e743c5320426b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af325efdca9ab36df4384e7c9f099bff8bb541817df94b7435eb06379dc124dd
+size 86880
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..586dbbe6e016563a28e12802739f9b4374841633
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a black shirt with white trim, standing in front of a red and white background. He appears to be speaking or listening intently, with his eyes closed and a serious expression on his face. The man is the main subject of the video, and he is positioned centrally in the frame. The background is simple and uncluttered, with a red and white color scheme that contrasts with the man's black shirt. The overall style of the video is straightforward and focused, with the man's actions and expressions being the primary focus."
+    ],
+    "video_ids": [
+        "KmyPkWjVmc8_16_47to251"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black shirt with white trim.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The man in the video is wearing a black polo shirt with white trim on the collar and sleeves, which matches the description. The video focuses on him, and no conflicting elements are present that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video in the video is wearing a black shirt shirt with white trim, the collar and sleeves, which matches the description provided The background does on the, and there additional elements are present.\"\n would the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..72f89510ea7ee3d8d1f92ec76f02559368777c86
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11a3f801150de59f829502d80376f67e93876539b9e58bb49038ab3245b07818
+size 113360
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..25b32744beebb3fd48382f04954492a451a1dda3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a taxidermy red panda in a museum setting. The panda is positioned in a way that it appears to be looking directly at the camera. The background consists of a wooden wall and a glass display case. The lighting in the scene is soft and warm, highlighting the red and black fur of the panda. The overall style of the video is realistic and detailed, capturing the lifelike appearance of the taxidermy specimen."
+    ],
+    "video_ids": [
+        "mWaI5GCuubQ_38_0to200"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A taxidermy red panda? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a detailed, lifelike representation of a red panda, which appears to be taxidermied or a highly realistic model, positioned in what looks like a museum exhibit. The object matches the description of a taxidermy red panda, with its distinct facial markings, fur texture, and posture. The surrounding elements (glass display case, other artifacts) do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a red and realisticelike depiction of a red panda, which align to be taxidermyied. highly highly realistic AI. given in a looks like a museum or. The fur in the description of a taxidermy red panda as as its fur fur features, fur texture, and posture. The setting environment,wood display case, museum exhibits in further not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..da39d5165dd2722798fcb0c54613c69aa76cf700
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a263dae53ab3a0b40effbaa8ceb26ab56c571ccce582c5960e56be3102076322
+size 118865
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..7b553c17df0cbc5b92ba1233405d47d800bf74b8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene outdoor scene featuring a light-colored dog, likely a Labrador Retriever, drinking water from a transparent plastic bottle placed on the grass. The dog is positioned centrally in the frame, with its head lowered towards the bottle. The background is softly blurred, suggesting a natural setting with greenery and possibly some trees or bushes. The lighting is warm and golden, indicating that the scene takes place during the late afternoon or early evening. The dog's fur appears smooth and well-groomed, and it has a calm demeanor as it drinks. The camera remains stationary throughout the sequence, focusing on the dog and the bottle, providing a clear view of the action without any noticeable movement or change in perspective. The overall atmosphere is peaceful and tranquil, emphasizing the simple act of the dog quenching its thirst in a natural environment."
+    ],
+    "video_ids": [
+        "21000118578326f5f5f657adfae3ccea852d91ae453273a3560222fa9a6d2bd9"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A light-colored dog (likely a Labrador Retriever) and a transparent plastic bottle on the grass.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a light-colored dog, consistent with a Labrador Retriever, drinking from a transparent container (likely a plastic bottle) placed on grass. The core elements described are present and accurately depicted, with no contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a light-colored dog, which with a Labrador Retriever, interacting from a transparent plastic onlikely a plastic bottle) on on the. The setting elements of in present and accurately depicted in fulfilling no additional or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..858ce7a670f55c5313dd2caa39c82f8a15c31945
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4709ba1e204f364da90e0265ee35a5773fce0c6e02d497d98fc748fe9d7c4160
+size 176613
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..811571384d7a8fb60693a1a778e47bb0e6807376
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a pair of black wireless earbuds in their open case. The case is black and has a soft, velvety texture. The earbuds are designed with a sleek, modern look, featuring a glossy finish and a minimalist design. The case is open, revealing the earbuds inside, and is placed on a white surface. The background is a wooden floor, providing a contrast to the white surface. The video is a simple, straightforward product shot, focusing on the earbuds and their case. The style of the video is clean and minimalistic, with a focus on the product."
+    ],
+    "video_ids": [
+        "cJIgkf_ND30_12_0to131"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Black wireless earbuds and a black earbud case.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black wireless earbud case with two black wireless earbuds inside, which matches the description. The case is open, and the earbuds are clearly visible in their designated slots. There are no conflicting elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a pair wireless earbud case with two ear ear earbuds inside. which matches the description of The case has open, revealing the earbuds are clearly visible, their designated slots. There are no additional elements in contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..de4a600566bc484e44e043509279b56b03a13f76
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:20e260b24b5f9a4b318e1ffc3be451a541bc3d3e345e25563362799dc5560b52
+size 167157
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..1d31a9e32a6737868cde152ecc3ab9bc9f2d139d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a car engine being worked on. The engine is open and exposed, revealing various components such as the air filter, spark plugs, and fuel lines. The engine is silver and black, and it appears to be in the process of being serviced or repaired. The style of the video is informative, with a focus on the details of the engine and the work being done. The video is likely intended for educational purposes, such as teaching viewers about car maintenance or engine repair."
+    ],
+    "video_ids": [
+        "bvez36pq1oc_23_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Car engine, air filter, spark plugs, fuel lines? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a car engine with visible components such as the air filter housing, spark plug wires (implied by the engine setup), and fuel lines (indicated by the various hoses and connections). Although spark plugs themselves are not individually visible, they are typically part of the engine assembly shown, and the video largely matches the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a car engine, various components such as the air filter,, spark plugs,,whichplied by the presence's), and fuel lines (imicated by the fuel tubes and tubes). The the plugs and are not directly visible, their are part part of the engine setup and. and the presence does matches the description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..69b210a4f2337df23b6216951fe4af00319059d6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-114-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d7b38eafd4fa71b312416bc018894f9e999d66c0c25b06b92b47cac93251490
+size 122519
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..e3f374e5654c74be8a79c73a062c3c1589470093
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up of a bowl of soup, showcasing the various ingredients and textures. The soup appears to be a hearty stew with a creamy base, containing chunks of meat, vegetables, and possibly some grains or legumes. The ingredients are arranged in a visually appealing manner, with the meat and vegetables being the most prominent. The soup is served in a white bowl, which contrasts with the darker colors of the ingredients. The lighting in the video is soft and warm, highlighting the textures and colors of the soup. The style of the video is simple and straightforward, focusing on the food without any additional context or background. The video is likely intended to showcase the dish's presentation and appeal to viewers' appetites."
+    ],
+    "video_ids": [
+        "0RSjudlFf6c_88_39to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bowl, soup, chunks of meat, vegetables, possibly grains or legumes? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white bowl filled with a thick, creamy soup-like substance that contains visible chunks of meat, various vegetables (like mushrooms and possibly onions), and what appear to be grains or legumes (such as beans or rice). The visual elements align well with the described object(s).\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close bowl containing with a brown, brown soup. substance. contains chunks chunks of meat, vegetables vegetables,possibly green and possibly green), and what appear to be grains or legumes.such as beans or peas). The presence elements match with with the description objects conditions).\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0d90ff0dd446ea3578b43a62427f7f2045ac2e8a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e1fe45d24cd228534dde7e8fe2a1f0951ee021173adc1847af59ab8931099315
+size 85653
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..0570bb96aeb55824b0ca8c5cff0724e783d53b45
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two men are standing at a table, each wearing a watch. The man on the left is holding a red smartphone, while the man on the right is holding a white smartphone. They are both looking at their watches, possibly checking the time. The table they are standing at is made of wood and has a few objects on it, including a plant and a book. The room they are in has a window, and there is a potted plant in the corner. The overall style of the video is casual and informal, with the focus being on the interaction between the two men and their smartphones."
+    ],
+    "video_ids": [
+        "-4s2MeUgduo_17_0to120"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, a red smartphone, a white smartphone, a plant, a book? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men standing at a table. A red smartphone is visible on the table, and both men are wearing smartwatches (one appears white, the other gray). A plant is visible in the background, and a book or framed picture is partially visible on a shelf behind them. All specified objects are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two men, at a table, One red smartphone and being in the left, and a men are holding wristwatches.which black to, the other black). There plant is visible in the background, and a book is stack object is on visible on the surface. the. The elements objects are present,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..313fcd1e95e09ffcf500912bcbbb70c9452d5d3a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c989c0c686ae2633cef22f2da4934e87052f5148b91322eb3529b6f9172eb2ba
+size 76961
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..cc842dccfb40665852c4239f4606151e21d81548
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a thrilling moment in a bicycle race. The main focus is a cyclist in a blue and yellow uniform, who is in the lead, leaning forward with determination. The cyclist's bike is prominently featured, with its sleek design and vibrant colors. The cyclist is surrounded by other cyclists, all in various stages of the race, creating a dynamic and competitive atmosphere. The road they are racing on is a dirt road, adding a rustic charm to the scene. In the background, there are trees and a car, suggesting that the race is taking place in a rural or semi-rural area. The overall style of the video is action-packed and full of energy, capturing the intensity of the race and the beauty of the surroundings."
+    ],
+    "video_ids": [
+        "WWVoFpEyh_I_36_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A cyclist in a blue and yellow uniform, other cyclists, a bicycle with a sleek design and vibrant colors, trees, a car.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a cyclist in a blue and yellow uniform, other cyclists in the background, a bicycle with a sleek design and vibrant colors, trees lining the road, and a car visible behind the cyclists. All elements described in the condition are present and do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a cyclist in a blue and yellow uniform, with cyclists, various background, a bicycle with a sleek design and vibrant colors, trees, the path, and a car in in the cyclists. The elements match in the condition are present in match not contradict the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c650125849d70a20ddff5c85df2eb7f2734f29dc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1accf540aaabb8186c0096bad4aad5dbf5b759d634bf8f836faaca77a02f3d23
+size 334768
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..79d48afedddeeefd9391770ecbc595b3876f3263
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing next to a silver car on a gravel road. The man is wearing a grey shirt and a baseball cap, and he appears to be in a good mood, smiling as he looks off to the side. The car is parked on the side of the road, and it has a license plate that reads \"B0D15\". The man is identified as \"BOB LUCIO, FORMER GOLF CLUB OWNER\", suggesting that he might be discussing his experiences in the golf industry. The setting seems to be a rural or semi-rural area, as there are trees and bushes in the background. The overall style of the video is casual and informal, with the man appearing to be relaxed and at ease."
+    ],
+    "video_ids": [
+        "BYT_n1_IdGY_21_0to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a silver car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man standing next to a silver car, which matches the 'Object(s)' condition. The man is identified as Bob Lucio, a former golf club owner, and the silver car is prominently featured in the background. There are no elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man standing next to a silver car, which matches the 'Object(s)' condition. The man is wearing by a,iano, and person professional player pro, and the car car is a displayed in the foreground. The are no additional in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..130702571da310a83dd3d1a85f50895b22192c3c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e61d72db61888b6b2d38bda93e50b5c8194d9903a3ef9709aad2f85e087f385
+size 173540
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..346c796410b81d28a84dfbcc9689280642db3dd2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young man with a beard and mustache, wearing a red baseball cap and a black and white checkered shirt. He is standing against a blue background with vertical stripes. The man appears to be speaking or gesturing with his hands, suggesting that he might be giving a presentation or explaining something. The style of the video is casual and informal, with a focus on the man's facial expressions and body language. The lighting is bright and even, highlighting the man's features and the details of his clothing. The background is simple and uncluttered, allowing the viewer to focus on the man and his message."
+    ],
+    "video_ids": [
+        "4YDrUlmIuDE_3_0to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man with a beard and mustache, wearing a red baseball cap and a black and white checkered shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man with a beard and mustache, wearing a red baseball cap and a black and white checkered shirt, which matches the description. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young man with a beard and mustache, wearing a red baseball cap and a black and white checkered shirt. which matches the description provided The background is the elements in not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bac9e4779b770348cb6ba940822affed4e29d085
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8133dfdfdd2373d91370b51bdf8f497d345dbc3b4c25595832948de1a4deedae
+size 242728
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..d8f15bb857df793cde0c536d7a687ec99f644714
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a mannequin dressed in a suit and tie, standing in front of a mountainous landscape. The mannequin is positioned in the center of the frame, facing the camera. The background consists of a vast mountain range with red and brown hues, suggesting a desert or arid environment. The lighting in the scene is bright, with the sun shining directly on the mannequin, casting a shadow on the ground. The overall style of the video is realistic, with a focus on the contrast between the mannequin and the natural landscape. The mannequin's pose and attire suggest a formal or professional setting, while the rugged terrain in the background adds a sense of adventure or exploration."
+    ],
+    "video_ids": [
+        "zAISPmxu7Tg_28_85to235"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A mannequin dressed in a suit and tie? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a figure dressed in a suit and tie, which matches the description of a mannequin in formal attire. The figure is headless, which is a common characteristic of the 'Slender Man' character, but this does not contradict the core description of a mannequin in a suit and tie. The background is a canyon, which is not part of the description but does not conflict with it.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a figure dressed in a suit and tie, which align the description of a mannequin dressed a attire. The figure is standingless, which is a distinctive characteristic of man SlSlender Man' character, but the does not contradict the ' description of the mannequin in a suit and tie. The background, a styl, which is an part of the description but does not conflict with the either}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d3e73b3889964946a02d8ec21c2dfd49584fac41
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90c6b03029f1637e7e89f32327d6cf71d510190efe3e514904f95012efcc084a
+size 153040
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..22b97228206b1ed7344934c6c49f0f036857ac95
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are seen preparing a meal outdoors. The man, dressed in a white outfit, is holding a large pot over a fire, while the woman, wearing a yellow sari, is stirring the contents of the pot. The fire is built on a small stone platform, and the cooking process is taking place in a field with trees in the background. The overall style of the video is simple and natural, capturing a moment of traditional cooking in an outdoor setting."
+    ],
+    "video_ids": [
+        "NZ4Fe0HyNTc_37_0to128"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a white outfit holding a large pot over a fire, a woman in a yellow sari stirring the contents of the pot.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a white outfit holding a large pot over a fire, and a woman in a yellow sari stirring the contents of the pot. These elements match the description provided, with no significant contradictions or inaccuracies.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a white outfit and a large pot over a fire, and a woman in a yellow sari stirring the contents of the pot. The elements match the description provided, indicating the additional additional or additionalacies.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5947631e5f04f59758fb52087099447e61804676
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9169ef8b15e2e5872dbf689d5e85825342c38af42dd22ec1112e5c6b95e1108a
+size 244346
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..ecafa02f963fb85f3b6aa9b82dd7d9b2ae2a7722
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a gray and white cat being wrapped in a beige towel. The cat appears to be comfortable and relaxed as it is being swaddled. The cat's eyes are open and it seems to be looking directly at the camera. The towel is being wrapped around the cat's body, covering it completely. The cat's ears are sticking out from the top of the towel, and its whiskers are visible. The cat's tail is not visible in the frame. The cat's fur is soft and fluffy, and it appears to be well-groomed. The towel is soft and plush, and it seems to be providing a warm and cozy environment for the cat. The cat's eyes are bright and alert, and it seems to be enjoying the attention it is receiving. The cat's mouth is closed, and it does not appear to be making any sounds. The cat's nose is visible, and it seems to be breathing normally. The cat's eyes are a light color, and they are looking directly at the camera. The cat's ears are perked up, and it seems to be listening attentively. The cat's whiskers are white, and they are sticking out from the sides of its face. The cat's tail is not visible in the frame. The cat's fur is a mix of gray and white, and it appears to be well-groomed. The towel is a"
+    ],
+    "video_ids": [
+        "8e0AWKlxgBU_14_1186to1360"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ['gray and white cat', 'beige towel']? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a gray and white cat wrapped in a beige towel, lying on a white surface. The cat's fur pattern and the towel's color match the description. There are no conflicting elements that contradict the specified objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a gray and white cat, in a beige towel. which on a surface surface. The cat's fur and and the towel's color match the description provided There are no additional elements in contradict the given objects.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e950519481f0c4f5a6b7049bed307bf6fb733351
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1140-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5569691fdd08494c00523a7f086c2f4faeec004793fc8bd8de54f976fa7b80b6
+size 343954
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..2a519256874d8df8c3cfde75ad981439e601cb2b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a cyclist navigating a narrow, cobblestone street in a European city. The cyclist, dressed in a black and red outfit, is seen leaning into a turn, skillfully maneuvering the bicycle on the uneven surface. The street is lined with parked cars and motorcycles, adding to the urban feel of the scene. The buildings on either side of the street are old and charming, with ivy climbing up their walls, adding a touch of nature to the urban landscape. The cobblestone street, the cyclist's dynamic pose, and the old buildings create a sense of movement and life in the city."
+    ],
+    "video_ids": [
+        "Gqws4xEZRSU_18_97to304"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A cyclist dressed in a black and red outfit, parked cars, motorcycles, old buildings with ivy.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a cyclist in a black and red outfit riding a bicycle on a cobblestone street. There are parked cars and a scooter (motorcycle) visible. The background features old buildings with ivy growing on the walls, matching the description. All core elements are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a cyclist dressed a black and red outfit riding a bicycle on a cobblestone street. There are parked cars and motorcycles motorcycle visiblewhichcycle) visible along The buildings features old buildings with ivy growing on them walls, which the description provided The elements elements of present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2d5d04d6ed3891560dfb3afa356ba0362307cae5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2d99038744c9c0c99fa57118701a513121c7499e5cc3995a274e7e12644fa798
+size 398234
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..258d478cb5acf6e6da33bfe73778be47bf26bd1a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit sitting at a table with a laptop and a plate of food. He appears to be in a formal setting, possibly a conference room or a meeting area, as indicated by the presence of a flag and a chair in the background. The man is engaged in an activity that involves the use of a red object, which he holds in his hands. The laptop is open and appears to be in use, suggesting that the man might be working or preparing for a presentation. The overall style of the video is professional and formal, with a focus on the man and his actions."
+    ],
+    "video_ids": [
+        "6sjkEVUl1ao_18_57to182"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a suit, laptop, plate of food, red object? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit sitting at a table with a laptop and a plate containing a red object, which matches the core description. Additional elements like flags and a grenade do not contradict the presence of these core objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit sitting at a desk with a laptop, a plate of food red object, which appears the description description. The elements like the in a notebook are not contradict the core of the core objects.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4de8703f81039da5ee0e5308247f986e7cfec5b3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:933112298feb7460ed7342f0e15baff4647eb9c18ea4463c22a57d956f7fb141
+size 183312
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..51bd6c99446a02c9ee748d9f24b12798ab3a506b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man with a beard and a yellow crown on his head is the main subject. He is standing in front of a black curtain, which serves as the backdrop for the scene. The man is wearing a purple shirt and a name tag that reads \"BARNS\". He appears to be in a room with other people, as there are two other individuals visible in the image. The man is gesturing with his hands, suggesting that he might be speaking or explaining something. The overall style of the video seems to be casual and informal, with the man in the crown being the focal point of the scene."
+    ],
+    "video_ids": [
+        "KoJd09yxK_I_21_0to134"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and a yellow crown, a man in a purple shirt with a name tag reading 'BARNS', another individual? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard wearing a yellow crown, a purple shirt under a light blue shirt with a name tag that reads 'BARNS'. Another individual's head is visible in the foreground, confirming the presence of at least one other person. The description matches the visual content without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and a yellow crown and and man shirt, which white-colored shirt, a name tag that reads 'BARNS'. There individual is shoulder is partially on the background, but the presence of another least two more person. The description align the video elements of any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cea13537016665dd789c22600a49c6296f020c7f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b36e60376cad57a73dd3c9ebc88a782fe45cb74a0b3d01a946d30375082886d
+size 143068
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..f9bb21e36fbaf841f58e502e403d70b48d148eb5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is an aerial view of a historic fortress perched on a cliff overlooking a city. The fortress is constructed of stone and features multiple levels and towers. The city below is densely populated with buildings of various sizes and colors, including red and orange rooftops. The surrounding landscape is lush with greenery, and the fortress is nestled among the trees. The sky is clear and blue, suggesting a sunny day. The video captures the grandeur of the fortress and the bustling city below, providing a unique perspective of the historic site and its surroundings."
+    ],
+    "video_ids": [
+        "yqKbaAOiMvM_31_0to152"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Historic fortress, multiple levels and towers, city buildings, greenery, trees, and the sky.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a historic fortress perched on a hill, featuring multiple levels and towers, surrounded by city buildings, greenery, trees, and the sky. The aerial perspective captures the fortress's architectural details, the surrounding landscape, and the distant city and sea, aligning well with the described elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a historic fortress withched on a hill, with multiple levels and towers. surrounded by city buildings with lushery, trees, and a sky. The architectural view shows the fortress's strategic details and the surrounding urban, and the urban horizonscape sea, alling well with the given elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dc9d4f839e222ba8b418c44eafdacbbf2a475f04
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d54d59f2c8c984b5d2c2ec23e039666fcf371d0254116f053a3a7a5b5a8b5766
+size 213064
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..d766de3a0fffa3108c6dbaae44e3048a07bceab2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a scene from a movie or TV show, featuring a young woman in a black jacket and green sweater. She is standing on a porch with luggage behind her, looking surprised or shocked. The porch is adorned with potted plants and a bench. The woman is holding a green bag and appears to be in conversation with someone off-camera. The scene is set in a residential area with a house in the background. The overall style of the video is realistic, capturing a moment of surprise or shock in the woman's life."
+    ],
+    "video_ids": [
+        "7-5kYQLJnFw_4_0to101"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman in a black jacket and green sweater, a green bag, and luggage.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The young woman in the video is clearly wearing a black jacket over a green sweater, and she has a green bag strap visible over her shoulder. In the background, there are also pieces of luggage (a green and a red suitcase), which aligns with the description. The core elements of the description are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video woman in the video is wearing wearing a black jacket and a green sweater, and she is a green bag in visible, her shoulder. The the background, there is hints hints of luggage,a black bag a black bag) which align with with the description of The video elements of the description are present represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7e0062a069fe7d321e9ff5f3e4ee323d8783a482
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2fd16aaecd660f2391c2fb5f5f70e5238385af51d120593da8582ba4da25d3fb
+size 187102
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..fdc689ab204b58ce4a8691080d3f54e4858e42c3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a man in a suit and tie, who appears to be a public official or a professional. He is seated at a desk with a microphone in front of him, suggesting that he is either giving a speech or participating in a press conference. The man is wearing glasses and is looking down at some papers, indicating that he is reading or reviewing information. The background shows other people, likely fellow officials or journalists, seated in a row of chairs, which implies that this is a formal event or meeting. The style of the video is professional and formal, with a focus on the man and his actions. The video likely includes a news ticker at the bottom of the screen, providing additional information or context about the event or the man's speech."
+    ],
+    "video_ids": [
+        "H2sNHYx7uJA_2_24to160"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, a microphone, papers, and other people seated in chairs.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a suit and tie seated at a table with a microphone in front of him. He is handling papers, and other people are seated in chairs in the background. All core elements described are present in the video without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a man in a suit and tie seated at a table with a microphone in front of him, He is holding papers, which there people are seated in chairs in the background. The elements elements of in present, the video, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ba191f9134405a88fbd61c6cdf8c16121628ebfb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:871a5756d768d5a68afde0586f2a0db16f1b492c9b1af9acd8cde249895f1b0c
+size 157469
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..b3689f44535918ce589eb5771f31b4aafed898ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a long beard and a baseball cap, standing in a garage or workshop. He is wearing a grey t-shirt with a logo on it. In the background, there is a white car with its hood open, and various tools and equipment are scattered around the workspace. The man appears to be a mechanic or a car enthusiast, and the video likely captures a moment of his work or hobby. The style of the video is casual and informal, with a focus on the man and his surroundings."
+    ],
+    "video_ids": [
+        "4cBKjbobwec_1_20to188"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a long beard and a baseball cap, wearing a grey t-shirt with a logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The man in the video has a long beard, is wearing a baseball cap, and a grey t-shirt with a visible logo that reads 'The Humble Mechanic'. These elements match the description exactly, and no contradictory elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video in the video has a long beard and is wearing a baseball cap, and is grey t-shirt with a logo logo. resembles 'V'ble'.ic'. The elements match the description provided, indicating there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b3fbcfa6eebb545d95eb13eb0315cc306810faab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50c505cbcb95a3ac79dd444c457a735943d76fdf58a261ede6a8ed7af7e4ebf9
+size 147008
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..9d3fd6bd9e33d71c7ae4725788f94a4baa1d69ea
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a long beard and a casual demeanor, standing in a forested area. He is wearing a gray t-shirt and appears to be gesturing with his hands as he speaks. The forest around him is lush and dense, with trees and foliage filling the background. The lighting suggests it might be late afternoon or early evening, with the sun casting a warm glow on the scene. The man's relaxed posture and the natural setting give the video a peaceful and serene atmosphere."
+    ],
+    "video_ids": [
+        "O66jIgc-dtI_28_0to172"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a long beard and a casual demeanor, wearing a gray t-shirt, gesturing with his hands as he speaks.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a long beard, wearing a gray t-shirt, standing in a forested area with a tent in the background. He is gesturing with his hands while speaking, which aligns with the described 'Object(s)' condition. The setting and his demeanor are casual, and no elements contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man with a long beard and wearing a gray t-shirt, and in a lushed area. sunlight natural in the background. He is gesturing with his hands as speaking, which aligns with the description 'cas(s)' condition. The setting and the casual match consistent, matching the additional contradict the description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fe53f3c6373b2f57c824554036dd4b74620dec94
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1141-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:49c10c27a834e59c037bfea5210430a1c2a7eda223547e992fc12546e248074f
+size 190686
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..bdc61b87ad98038363f36b28f9e8bc9b4706e11b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and green eyes, wearing a white t-shirt with a black graphic on it. He is holding a small object in his hand, which appears to be a pen or a similar item. The man is standing in a room with a green shower curtain in the background. The lighting in the room is dim, creating a somewhat moody atmosphere. The man's facial expression is one of surprise or shock, as he looks directly at the camera. The overall style of the video is casual and candid, with a focus on the man's reaction to the object he is holding."
+    ],
+    "video_ids": [
+        "7F-8ZpvN30Q_17_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and green eyes, wearing a white t-shirt with a black graphic. He is holding a small object, possibly a pen.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and green eyes wearing a white t-shirt with a black graphic (AC/DC logo visible). He is holding a small object, which appears to be a Wahl Peanut trimmer, not a pen. The core description is accurately fulfilled, and the object held is consistent with grooming tools, which is plausible in context.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and green eyes, a white t-shirt with a black graphic.likelyDCDC logo).). He is holding a small object, which appears to be a pen clip Trmer, matching a pen. The description elements of largely represented, with the additional in by consistent with the tools, which is a for a.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..13180f81d4094368419c0f7542a36c9475f097d1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:475d479609af201be0470d7ec858f738497a5b0a21a47467a42a1a9f92bc76ba
+size 137326
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..d2e1119185efef0418a6da63fe924c7dbaec5261
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen in a store, holding a drink in his hand. He is wiping his face with his hand, possibly after taking a sip. The store has a festive atmosphere, with a Santa Claus figure visible in the background. The man is wearing a green shirt, and the store has a white ceiling with lights. The overall style of the video is casual and candid, capturing a moment in the man's day."
+    ],
+    "video_ids": [
+        "O-C6aDuerV8_12_0to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a drink, a Santa Claus figure? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man holding and drinking from a glass, and a Santa Claus figure is visible in the background. These elements match the 'Object(s)' condition described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man holding a drinking from a cup, which there Santa Claus figure is visible in the background. The elements match the 'Object(s)' condition described in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..859190b69c6af109ab5edca31b7ab03898bec024
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:580a13d09ce7780f1e44eadef72c3801a5db12ca0f07136870a1300e68046bfa
+size 189444
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..14323b9e303126b4180d4a1030c5fa2c3096b6b8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring three men in suits sitting at a desk with a sports theme. The men are engaged in a discussion, with one man speaking while the other two listen attentively. The desk has a red and black color scheme, and there are sports-related graphics on the screen behind them. The graphics include logos and text related to a sports event. The men are dressed professionally, with one man wearing a blue suit, another in a black suit, and the third in a gray suit. The overall style of the video is professional and polished, with a focus on sports commentary."
+    ],
+    "video_ids": [
+        "ZIUwKoIUB4k_8_0to175"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three men in suits, a desk, sports-related graphics on a screen? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three men dressed in suits sitting at a desk, which matches the core description. Additionally, there is a screen displaying sports-related graphics for the '2010 BCS National Championship' with team logos, fulfilling the 'sports-related graphics on a screen' condition. No elements contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows two men in in suits, at a desk, which align the description description of The, there are a sports in sports-related graphics in the backgroundTO021 E' Championship Championship' in a logos, which the 'sports-related graphics on a screen' condition. The additional contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..189bb91337606acd9070e38e79bb263a6bf69cb4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e9ab6d85eb991e2404c0b42e0f5dcd627e7d3a17e5c92e818548f67f1d15c86
+size 82292
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..a42c994055d0122ec8befaf3edbecfd89ea83c4a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a police officer in a uniform and sunglasses is seen interacting with two individuals in a desert-like setting. The officer is standing in front of a large black truck with a logo on the side. The first individual, wearing a camouflage hat, is holding a piece of paper and appears to be speaking with the officer. The second individual, also wearing a camouflage hat, is standing nearby, observing the interaction. The scene suggests a possible traffic stop or a routine check in a rural area. The style of the video is realistic, capturing a moment of interaction between law enforcement and civilians in an outdoor setting."
+    ],
+    "video_ids": [
+        "36Sk81nr9kI_5_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Police officer in uniform and sunglasses, two individuals wearing camouflage hats, a large black truck with a logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a police officer in uniform and sunglasses, two individuals wearing camouflage hats, and a large black truck with a logo (Arizona Game and Fish Department). All elements described in the condition are present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a police officer in uniform and sunglasses, two individuals wearing camouflage hats, and a large black truck with a logo.part). and Fish Department). The elements in in the condition are present in match depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..50bc1baec586bf6989f88031e231c9b1169fa1a5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:05d3f21e968a7fd41fbf9890a73a1818ceffb711760d1ef573b1616af7435eb7
+size 178822
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..c4413ee6c404cf7582b5bf03d278b6a89d7d0502
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a basketball player in action during a game. The player, wearing a white jersey with blue and orange accents, is seen in three different frames. In the first frame, the player is seen looking up, possibly tracking the ball or preparing for a play. In the second frame, the player is seen in motion, running towards the basket, with a focused expression on his face. In the third frame, the player is seen in a defensive stance, arms outstretched, ready to intercept the ball. The player's jersey is clearly visible, with the team's logo prominently displayed. The background of the video is a blur of the basketball court, indicating the fast-paced nature of the game. The player's movements and expressions suggest a high level of concentration and athleticism."
+    ],
+    "video_ids": [
+        "FvIFcwpZKJs_2_0to161"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A basketball player in a white jersey with blue and orange accents.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a basketball player wearing a white jersey with blue and orange accents, which matches the description. The jersey has 'THUNDER' written on it, and the number '7' is visible, indicating it is a team uniform. The player is also wearing a white headband and arm sleeves, which are additional elements but do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a basketball player wearing a white jersey with blue and orange accents, which matches the description. The player features theLOUNDER' written on it, and the player '1' is visible on indicating it is a Thunder uniform. The player is holding holding a wrist wristband, wrist sleeves, which are typical elements that do not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..562934f99694400ba817c3c1d89d09f630c8d6ea
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6f12f08451f38c25bd274b2420db58638c5fef23d6d4e87050d494e8945b3d7
+size 273038
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..e1b29f9502e2feb24e151863332ee4d85b3bb88a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a plaid shirt standing in front of a television screen displaying the logo of \"D News\". He is gesturing with his hands, possibly explaining or discussing a topic. The background includes shelves with various objects, suggesting a setting that could be a news studio or a classroom. The man appears to be the main subject of the video, and his actions and expressions are likely the focus of the content. The overall style of the video seems to be informative or educational, with the man possibly presenting or teaching about a particular subject."
+    ],
+    "video_ids": [
+        "MI9oOHrVI5M_0_0to128"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a plaid shirt, a television screen displaying 'D News', and shelves with various objects.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a plaid shirt, a television screen displaying the 'D News' logo, and shelves in the background with various objects such as a microscope, molecular models, and framed butterflies. All core elements described are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man wearing a plaid shirt, standing television screen displaying ' textD News' logo, and shelves with the background with various objects. as books red and books models, and other pictures. These these elements of in present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..014f14437d34f76f986501bfa57e0ac94a9fba8e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a89ee52bee9ab85406a49140b4be5de100c98b6eeef37a6e1dda855eb43089fb
+size 215395
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..22b78e803533b3f2843b55a805e88ffa0c07aae8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the iconic Hollywood Pictures sign, a symbol of the film industry. The sign is set against a clear blue sky, with palm trees swaying gently in the background. The sign itself is a large, ornate structure, featuring two statues of elephants on either side. The elephants are intricately detailed, adding a touch of grandeur to the scene. The sign reads \"Hollywood Pictures\" in bold, capital letters, with the word \"Pictures\" written in a smaller font size. The overall style of the video is a blend of architectural and natural elements, creating a visually striking contrast. The Hollywood Pictures sign, with its rich history and cultural significance, serves as the focal point of the video."
+    ],
+    "video_ids": [
+        "P2lFPowLVaw_13_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Hollywood Pictures sign, two statues of elephants, palm trees? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the Hollywood Pictures sign, two statues of elephants atop ornate pillars, and palm trees in the background against a blue sky. All elements described in the condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a Hollywood Pictures sign, two statues of elephants, aate structures, and palm trees in the background. a clear sky. These elements match in the condition are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..abd95dae5c4dfabbee387e16311723faba7d61a4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64b995f1a2d29e96e1684c5e99281a79db20a451dc3209f9940d46e20da41efb
+size 170153
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..c3c3634a710b2136e2ced9c6a60b2f52eafaf7b0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a delicious meal being prepared. The main focus is on a piece of chicken, which is being cooked to perfection. The chicken is golden brown, indicating that it has been cooked until it is crispy on the outside. The chicken is garnished with a sprig of rosemary, adding a touch of green to the otherwise golden dish. The rosemary is placed on top of the chicken, suggesting that it has been used to season the dish. The background is blurred, drawing the viewer's attention to the chicken and rosemary. The overall style of the video is simple and straightforward, focusing on the food and the cooking process. The video does not contain any text or additional elements, keeping the viewer's attention solely on the meal being prepared."
+    ],
+    "video_ids": [
+        "b--OO1GnsFQ_31_18to153"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A piece of chicken, a sprig of rosemary? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a dish that appears to be a cheesy baked item, which could reasonably be interpreted as chicken (given the texture and context). A sprig of rosemary is clearly visible as a garnish on top. The presence of cherry tomatoes in the background does not contradict the core description, as they are additional elements that do not conflict with the primary objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a piece that includes to be a piece, item, possibly could be be interpreted as a,though the context and color). It sprig of rosemary is clearly visible on a garnish on top of The presence of additional tomatoes and the background does not contradict the core description but as they are not elements that do not conflict with the main objects mentioned}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..17290dffd24b334d26f0ebc8c69443f0830d6bcd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1142-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae49d73b2634bb5cdce1f1ade5d89d72247fcd13d762943446df319eaadb71ed
+size 122933
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..73d25009235e411e9145b8d216df6fa2a007f2cc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a plate of pasta, which appears to be fettuccine alfredo, being served. The pasta is creamy and yellow, with a generous amount of grated cheese sprinkled on top. The cheese is white and appears to be freshly grated. The pasta is garnished with a sprig of green herbs, which adds a touch of color to the dish. The plate is white and round, and it is placed on a table. The table is covered with a white tablecloth, and there are other dishes and utensils visible in the background. The lighting in the video is soft and warm, highlighting the creamy texture of the pasta and the golden color of the cheese. The style of the video is simple and straightforward, focusing on the food and the presentation of the dish."
+    ],
+    "video_ids": [
+        "Av0ULngGQA4_31_0to161"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A plate of fettuccine alfredo, grated cheese, and a sprig of green herbs.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a plate of fettuccine alfredo, identifiable by the wide noodles in a creamy sauce, topped with grated cheese and a sprig of green herbs (likely basil). Additional elements like a glass of beer, bread, and another plate in the background do not contradict the core description and are acceptable as contextual elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a plate of fettuccine alfredo, with by the wide, coated a creamy sauce. topped with grated cheese and garn sprig of green herbs,likely parsley or The elements such the fork of water and a, and a plate of the background do not contradict the core description and are acceptable as they elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1ad28f32b27c03c188b4ff710345a1531a28f7b2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e2e6d07f30dd242a2d7bdd9b40c9621ac4b7da167ed486b4911452968be09633
+size 180356
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..a03ce14ba423d7f8cb00e2fdf838622e25a47422
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a smiling expression, standing in front of an American flag. He is wearing a dark suit and a red tie. The man appears to be in a formal setting, possibly a press conference or a political event. The American flag is prominently displayed in the background, suggesting a patriotic theme or context. The man's smile and the flag's presence convey a sense of confidence and national pride. The style of the video is straightforward and professional, with a focus on the man and the flag."
+    ],
+    "video_ids": [
+        "UkyQxFCTUZI_16_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a smiling expression, wearing a dark suit and red tie.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a smiling expression, wearing a dark suit and red tie, which matches the description. The presence of the American flag in the background does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a smiling expression, wearing a dark suit and a tie. which align the description provided The background of an American flag in the background does not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..df8c1bf86acfa8775e9259a507c3e7e904cf5263
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2ecb75aa834cfddebbd6d0796b24ced795ae79cd0520e74e24e5e4845c90aa25
+size 101356
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..e02199695fce83d1916f626a13dd21b2281875ce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a woman with blonde hair sitting at a table with a white sewing machine. She is wearing a striped shirt and has a smile on her face. She is holding a spool of thread in her hand and appears to be in the process of sewing. The table is covered with various colorful fabrics and buttons. The background features a white wall with yellow polka dots. The style of the video is bright and cheerful, with a focus on the woman's enjoyment of sewing."
+    ],
+    "video_ids": [
+        "BR4t23BCYJ0_41_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a white sewing machine, a spool of thread, colorful fabrics, buttons? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman sitting at a table with a white sewing machine in front of her. She is handling colorful fabric pieces (likely felt or paper circles) and appears to be sewing them. Although a spool of thread is not clearly visible, the context implies it's part of the sewing setup. The colorful fabrics and buttons are represented by the various colored circular pieces scattered on the table. The scene matches the described objects without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman sitting at a table with a white sewing machine in front of her. She is holding a fabrics pieces andwhich the) similar)) and there to be engaged them. The there spool of thread is not visible visible, the presence of its is present of the sewing setup. The presence fabrics and buttons are also by the fabric shapes felt pieces and on the table, The scene align the description elements without any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1a449a31d791f2290a73524d701e1c5cc85c1188
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a4def8449cebd22bb55b46cd0e484312bd54ab3ae2c3a6c272c1d391f2717879
+size 148092
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..ad6236d2ca614929791b12ea8110077ad84b690e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is a player from the Atlanta Falcons, wearing a red and white uniform with the number 18. He is in the midst of a powerful run, holding the football securely in his hands. His helmet is black, matching the color of his uniform, and he is wearing white gloves. The player is running towards the right side of the frame, his body leaning forward in anticipation of the next play. The background is a blur of green, indicating the field, and there are other players visible, adding to the sense of action and competition. The style of the video is a fast-paced, action-packed sports sequence, capturing the intensity and excitement of the game."
+    ],
+    "video_ids": [
+        "22QGAnAsmsg_63_23to159"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player from the Atlanta Falcons (wearing red and white uniform with number 18), other players, and the football.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in an Atlanta Falcons uniform (red and white) with the number 18, holding a football and running. The player's gear, including helmet and gloves, is consistent with NFL standards. Although the background is blurred, it does not contradict the core description, and no other players are clearly visible, which is acceptable as the description does not require them to be present. The video successfully fulfills the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a player in a Atlanta Falcons uniform,red and white with with the number 18, holding a football and running on The player's uniform and including the and gloves, is consistent with the standards. The the player is not, it appears not contradict the description description of as the other players are visible visible in which is acceptable given the focus does not specify the to be present.\"\n The football successfully fulfills the 'Object(s)' condition by}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4303c9b902c9f972a92d59fae8a42a13232f1802
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d5ac0c6df994ca23df22dea630110c149c7e19fa797b58ac78c9ea4f3482918
+size 300269
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..072ab00ea0c9f49b5a0d17a7accb5c0a1729c3f7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen in a kitchen, making a heart shape with his hands. He is wearing a blue shirt and a red hat. The kitchen is equipped with a refrigerator, an oven, and a sink. The man is standing in front of the refrigerator, which is located on the right side of the frame. The oven is situated on the left side of the frame, while the sink is located in the middle. The man's action of making a heart shape with his hands suggests that he might be preparing to cook or bake something. The overall atmosphere of the video is casual and homely."
+    ],
+    "video_ids": [
+        "IZMLV4nEWsU_13_408to569"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a refrigerator, an oven, a sink, a blue shirt, and a red hat.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a blue shirt and a red hat (used as a blindfold), standing in a kitchen. A refrigerator and an oven are visible in the background. While a sink is not prominently visible, it is implied to be present in the typical kitchen setup. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing a blue shirt and a red hat,likely as a capfold). standing in a kitchen with The refrigerator, an oven are visible in the background, The the sink is not directly featured in it is implied to be part as the kitchen kitchen setting. The the elements of in present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1fc83b13f6bcf05b5cf7b7ee41e5392d4fd40f73
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb80ceb82e7e01d09101593e585d15f5c484a4e4281b996d9c80133107ca346d
+size 123101
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..b71aca8cccda7362412eb263085263ffcbef49d3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man sitting on a stool in a bustling market, engrossed in his work on a calculator. He is dressed in a white shirt and a blue apron, indicating his role as a vendor. The market is filled with various items, including books and boxes, suggesting a diverse range of products available for sale. The man's focused demeanor and the lively atmosphere of the market create a dynamic scene that encapsulates the essence of daily life in a bustling marketplace."
+    ],
+    "video_ids": [
+        "0dAuHLEXLyM_162_258to391"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a stool, a calculator? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man sitting on a stool while holding a calculator. These three objects are prominently featured and accurately match the 'Object(s)' condition described. Additional elements in the background (e.g., other people, shelves, boxes) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man sitting on a stool, using a calculator. The elements elements ( present featured in match represent the descriptionObject(s)' condition described. The elements like the video,books.g., books books, books, and) do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..334f419cefcfd755773a330e0ad0554f4536f869
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c017de2dbc89fa3d12c218506a8087913063c09295810f96c8e6caca23e3c77c
+size 204038
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..d538f535d47f35f99445940d2b7a711994d59521
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a kitchen, holding a plate of food. She is standing in front of a counter, which has a bowl and a vase on it. The kitchen is equipped with modern appliances, including a microwave and an oven. The woman is dressed in a black jacket and a floral dress. The video captures the woman's journey from preparing the food to serving it on the plate. The overall style of the video is realistic and it captures the everyday life of a person in a kitchen."
+    ],
+    "video_ids": [
+        "ySCSeITKRzk_11_0to142"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a plate of food, a counter, a bowl, a vase, a black jacket, and a floral dress.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a kitchen holding a plate of food. A counter is visible with a bowl and a vase on it. The woman is wearing a black jacket over a floral dress. All specified objects are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a woman in a kitchen setting a plate of food. The counter is visible with a bowl of a glass on it. The woman is wearing a black jacket over a floral dress. The the objects are present and the with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3a063de63dcfa8e0eb339281ed3964d045ade2cb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f22ec955bd5b4849ff4f698bb370cd5a44425d67636872685bbefd40e2fc9b11
+size 182754
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..31465b9f1a302be185cfb844cd86b0165e014540
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person's hand holding a device with a screen and buttons. The device appears to be a GPS or navigation system, given its design and the presence of a screen. The hand is pressing a button on the device, which is likely a red button, as indicated by the image. The device is held in front of a body of water, which could be an ocean or a large lake, given the vast expanse of water visible in the background. The style of the video is a close-up shot, focusing on the interaction between the hand and the device, with the natural scenery serving as a backdrop. The video captures a moment of navigation or exploration, possibly on a boat or near the water's edge."
+    ],
+    "video_ids": [
+        "DFsZwRVq2Ag_50_0to130"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A hand, a device with a screen and buttons (likely a GPS or navigation system), a red button on the device.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a hand interacting with a device that has a screen and multiple buttons, including a prominent red button labeled 'STANDBY'. The device appears to be a marine GPS or navigation system, as indicated by its design and the context of being on a boat with ocean in the background. All core elements described are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a hand interacting with a device that has a screen and buttons buttons, which a red red button. 'STOPANDBY'. The device appears to be a GPS GPS or navigation system, given indicated by the design and the context of being held a boat. a in the background. The elements elements of in present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3c362780823f6d94324142b6b82a5b3f7fd219af
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1143-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e2a262b36cb7ffb1a262c068ae5051c3c149b764f7924d0a9eee23b8ede9219
+size 144665
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..9f3237e9c0933c924962bf2e8391a54224945e82
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a black and white time-lapse of a man driving a car. The man is wearing a suit and tie, and he is seen in various positions as he drives. The car is a luxury vehicle with a steering wheel and dashboard visible. The man's hands are on the steering wheel, and he appears to be focused on the road. The car is moving, and the man's position changes as he navigates the vehicle. The time-lapse effect creates a sense of motion and progression. The black and white color scheme adds a timeless quality to the video."
+    ],
+    "video_ids": [
+        "adh6s4AKRVg_12_47to198"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man, car (luxury vehicle).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a man seated inside a luxury car, specifically a Bentley, which matches the description. The man is visible in the driver's seat, gesturing as if speaking, and the interior of the car, including the steering wheel, dashboard, and seats, is prominently displayed, confirming it is a luxury vehicle.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a man seated in a luxury vehicle. which a vehicle, as is the description of The man is dressed wearing the driver's seat, anduring with if he, and the interior of the car, including the steering wheel and dashboard, and seats, is prominently displayed. indicating the is a luxury vehicle.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..520671758f6ac717c79f346118611589ccbab247
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d68a8e27bb1d9ea7a86c972b3404f2bf605baa614eb6835332142db64a199593
+size 257068
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..a289d482a3b8207c63a074175b2cc216effd397b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit standing in front of a building with a green light. He is smiling and looking to his left. The camera is positioned to the left of the man, capturing him from a side angle. The man is wearing a gray suit and a black tie. The building behind him has a green light on it. The man appears to be posing for the camera. The overall style of the video is a candid, real-life moment captured on camera."
+    ],
+    "video_ids": [
+        "12HFHhJfDEM_44_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a gray suit with a black tie, a building with a green light.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a man in a gray suit with a black tie, as described. In the background, there is a building with a visible green light, matching the description. While there are additional elements like photographers and bystanders, they do not contradict the core description and are consistent with a public event setting.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a man dressed a gray suit with a black tie, standing described. The the background, there is a building with a green green light, which the description. The the may no elements in the and aers, they do not contradict the core description and are acceptable with the public event or.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c8a6f3112d22c607dbad9f5a537300c1838b700a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2cf9d00c0e446375ffd24cf2fd7493eff5ca02eae04f0d2c323b26fe35d87ef7
+size 97627
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..113e8537004fdfc33fb491433971f4dd92ca8a4d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a pond teeming with koi fish. The water is clear enough to reveal the sandy bottom beneath, dotted with small rocks and patches of aquatic plants. A variety of koi fish, ranging in color from vibrant yellows and oranges to deep reds and blacks, swim gracefully through the water. Some fish are larger, while others are smaller, creating a dynamic and lively movement throughout the frame. The fish move in different directions, some swimming close together, while others glide alone, occasionally creating ripples on the surface. The overall atmosphere is peaceful, with the gentle motion of the fish providing a soothing visual experience. The camera remains stationary, allowing viewers to fully absorb the tranquil beauty of the koi pond."
+    ],
+    "video_ids": [
+        "4972ffd52632f9da058c0eecfc50c3cc3fb9e75a269016b5828af80c758b1218"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Koi fish of various colors including yellow, orange, red, and black.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group of koi fish swimming in clear water, displaying a variety of colors including yellow, orange, red, and black. These colors are clearly visible across multiple fish in the frame, fulfilling the specified condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a variety of fishoi fish in in a water. with a variety of colors including yellow, orange, red, and black. The colors are consistent visible and the fish, the scene, fulfilling the ' ' of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..69fbb12eb110b4eb48d3df79deab6aad5bdcc7fd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c2c7517f688d0bc5d07897c25368dd3bb663df373269e3902885aa42f838799a
+size 277019
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..df54eee8158de38beec5617c614456d669899ebc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit and bow tie, sitting in a chair and speaking. The man appears to be in a professional setting, possibly an office or a studio. He is wearing a suit jacket and a bow tie, suggesting a formal or business-like environment. The man is speaking, which indicates that he might be giving a presentation or an interview. The background is plain and white, which puts the focus on the man and his speech. The style of the video is straightforward and professional, with no additional elements or distractions. The man's attire and the setting suggest that the video is likely related to a business or professional context."
+    ],
+    "video_ids": [
+        "Cv8AnOgkThU_7_469to600"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and bow tie sitting in a chair and speaking.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a suit jacket and a bow tie, seated in a chair, and actively speaking. These elements align with the core description provided, even though the background is plain and there are no additional conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed a suit and, a bow tie, seated in a chair, and appears speaking. The elements align with the description description provided. indicating though the video is not and there are no additional elements elements.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..da4cce87f5839615573f52f3d9579e43f3e7fb40
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17c119db710cc39f546c77cadbe167783bd4dc0a6bd5e1a5282d0a0201930714
+size 87749
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..574d8d6dad1ae7bf1718e1eca3118c4c53c89c2f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a television interview featuring two men in suits sitting in chairs facing each other. The man on the left is older with white hair, wearing a dark suit and tie, and appears to be speaking. The man on the right is younger, wearing a light-colored suit and tie, and is listening attentively. Between them is a small round table with a couple of microphones on it. The background is a studio setting with a large screen displaying a clock and a graphic with the text \"Mes PASO 2019\" and \"409 48 13\". The style of the video is a professional news or talk show format, with a focus on the conversation between the two men."
+    ],
+    "video_ids": [
+        "7rYquIsBG28_20_1253to1425"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men in suits, a small round table, a couple of microphones.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men in suits seated in armchairs, with a small round table between them. Microphones are visible on the table, and the setting is consistent with a television interview or talk show. Additional elements like a countdown timer and text overlays do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men in suits seated across achairs, facing a small round table in them. Onphones are placed on the table, and the background appears consistent with a studio studio or talk show. The elements such the clock clock and a in do not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1e3e756b52807850d580866522a7e2b86c1f4649
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c967679c4557fda74d5235426cc579ef35a9fdc90f01782cfffd2637630c543a
+size 66004
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..89dfe86ba4689af511344db63ecbe94f8a3e3e91
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man named Geoffrey A. Fowler, who is a personal technology columnist. He is seen in a professional setting, likely a news studio, with a cityscape in the background. Geoffrey is wearing glasses and a light blue shirt, and he appears to be speaking or presenting. The video is likely a news segment or an interview, as indicated by the presence of the cityscape and the professional attire of the man. The style of the video is informative and professional, with a focus on the man and his expertise in personal technology."
+    ],
+    "video_ids": [
+        "RE8-Pgb2Doo_1_0to185"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Geoffrey A. Fowler (a man), glasses, light blue shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing glasses and a light blue shirt, who is identified as Geoffrey A. Fowler, Personal Technology Columnist. The visual elements match the description accurately.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man wearing glasses and a light blue shirt, which appears identified as Geoffrey A. Fowler. as Finance Editorist for The presence elements in the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..62592644685c7748be2fe22406cd85ebdf7f07dc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f218666f0dc29cf22e484b009372a03d20514e85c0281e0e8369f5c3b4cd8f09
+size 226570
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..91c9fb20c662988334eda01befeaedea5dff5667
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a screenshot from a video game, specifically a football (soccer) match. The game is paused, and the user interface is visible. The scoreboard shows \"SL BENFICA 0 - 1 SPORTING CP\" with the time \"15 February 2020 16:00\" and the league \"Liga NOS\". The user interface includes options for \"Resume\", \"Instant Replay\", \"Trainer\", \"Match Facts\", \"Settings\", \"Forfeit Match\", and \"Exit\". The player's team formation is displayed on the left side of the screen, with player names and positions. The player's avatar is visible in the bottom left corner, wearing glasses and a yellow shirt. The background shows a football stadium with a green field and red stands."
+    ],
+    "video_ids": [
+        "BFgH7Dy7QlA_11_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Scoreboard, User Interface, Player's Team Formation, Player's Avatar? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly displaying a scoreboard (showing SL Benfica 0-1 Sporting CP), a user interface (with options like Resume Match, Instant Replay, etc.), a player's team formation (SL Benfica 3-4-3 Diamond layout with player names and positions), and a player's avatar (the man in the orange shirt in the bottom left corner). All these elements are present and clearly visible.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video contains fulfills the 'Object(s)' condition as displaying showing a scoreboard attoping 'IDEfica and0 and00 CP), a user interface withwith various like ',, Save Replay, and.), a player's team formation (with Benfica's4-4-3),),), player positions and positions), and a player's avatar (a character with the bottom shirt). the bottom right corner). The these elements are present and match visible in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d7ff8b46f406dacf451be0c69696658f042dc157
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8da023d65f4d38bf7cb66acd2cbe9b3b4bae337f3bb04b4611bd9c67c755a594
+size 115531
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..dbe81c1f01623d2923383cf5fbfd920275b400e4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a playful and colorful depiction of a beach scene with a group of toy figurines. The figurines are dressed in summer clothing, with some wearing swimsuits and others in casual beachwear. They are standing on a sandy beach, with a few of them interacting with each other. The beach is adorned with a few colorful items, including a bucket and a shovel, suggesting that the figurines are engaged in a fun beach day. The overall style of the video is whimsical and lighthearted, with a focus on the joy of a day at the beach."
+    ],
+    "video_ids": [
+        "5j2D6I2i65E_13_237to366"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A group of toy figurines dressed in summer clothing, some in swimsuits and others in casual beachwear, standing and interacting with each other.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group of toy figurines on a sandy surface, dressed in summer clothing such as swimsuits and casual beachwear. Some are in swimsuits (e.g., the boy in the yellow tank top and blue shorts), while others wear casual beach attire (e.g., the girl with pink hair in a green dress). They are positioned as if standing and interacting with each other, with one doll being moved by a hand, suggesting interaction. The presence of other small objects like a surfboard and a cupcake does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a group of toy figurines dressed a sandy surface, which in summer clothing, as swimsuits and casual beachwear. They figur in swimsuits,pink.g., pink two in the foreground and top and blue shorts, while others are casual beachwear (e.g., the man in the shorts in a pink top and The are standing as if they and interacting with each other, which some figur holding held slightly a hand, suggesting a. The presence of beach beach objects like a blueboard and a bucket adds adds not contradict the core description but}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3aa41dd4398609900c23b046e004d60a6c3defdd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1144-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c9ca0134666851df60c8f69eeace37ee9d868cab7aa93a6c48f617236d389c6
+size 125315
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..587257322f8dfd5a829b481d5b594cb7cc79b0ff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a journey on a winding mountain road. The road, a ribbon of gray, twists and turns through the verdant landscape, with a stone wall on one side and a steep drop-off on the other. The car, a small red vehicle, is seen driving along the road, its vibrant color contrasting with the surrounding greenery. The road is flanked by lush green mountains, their slopes dotted with patches of snow. The sky above is a clear blue, with a few clouds scattered across it. The overall style of the video is a dynamic aerial shot, providing a bird's eye view of the journey. The camera follows the car as it navigates the winding road, capturing the beauty of the mountainous landscape and the thrill of the drive."
+    ],
+    "video_ids": [
+        "nR5wzEnV6qo_50_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small red car driving on the road.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a small red car driving on a winding mountain road, which matches the core description. The surrounding scenery, including snow patches and green hills, does not contradict this description and enhances the context.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a small red car driving on a winding road road. which matches the description description of The presence environment, including the-c on greenery, adds not contradict the description and adds the visual of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fe2e1b3dce267763b13c8ee56f0956e5941b2f25
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5db2aa3ceabfbd1e55995016396267747ade2c22cf221d075657da73de654d40
+size 336553
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..f5f3df7786dc7954e8be11ddc56b2044f8713829
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a classic car, focusing on the driver's seat and dashboard. The car has a vintage design with a blue interior, and the dashboard features a variety of gauges and controls. The car is parked inside a building, as evidenced by the reflections in the side mirrors and the visible interior of the building. The style of the video is straightforward and documentary, capturing the details of the car's interior without any additional narrative or embellishment. The video is likely intended for a car enthusiast or collector, showcasing the classic design and features of the vehicle."
+    ],
+    "video_ids": [
+        "_IGudSd0Hks_5_0to104"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Classic car, driver's seat, dashboard, gauges, controls? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the interior of a classic car, including the driver's seats, dashboard, gauges, and controls. The camera pans from the front seats to the dashboard, highlighting these elements. Although the background includes neon signs and a showroom setting, these do not contradict the core description of the car's interior components.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts the interior of a classic car, including the driver's seat, dashboard, gauges, and controls. The presence angle across the driver to to the dashboard, highlighting the elements. The the video is a signs and a building-like, which do not contradict the core description of the classic's interior..\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1035f3084491e1aa6f8c56eda1f8866fafb1475e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63cee961a4d7ca49b4d4d564514d8348dd48d8aa8a7a1b760a9be198787fd044
+size 230387
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..f15dc7b51195c4c161e7e73ed39dff96dd732e5b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a brown bear and a gray wolf are depicted in a friendly encounter. The bear, wearing glasses, stands on a rocky outcropping, while the wolf stands on a cliff above. The background features a clear blue sky with fluffy white clouds. The scene is set in a natural environment, possibly a forest or mountainous area. The interaction between the two animals suggests a peaceful coexistence, with no signs of aggression or fear. The overall style of the video is cartoonish, with vibrant colors and exaggerated features that give the animals a friendly and approachable appearance."
+    ],
+    "video_ids": [
+        "YdF9OEWruJA_9_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A brown bear and a gray wolf? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a gray wolf standing on a rock and a brown bear looking up at it, matching the described objects. The background and additional elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a brown wolf and on one cliff, a brown bear standing at at the. which the description '. The presence and additional elements do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..13e62737c3da25ada3df08cabca4968213c86ea3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0461ba653d208a1b51d88456c67ef5f264560432e92fd90dcca08807ce368505
+size 184275
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..05e9e51f65e6f517169578dbba8788e5755afb5c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a white sports car parked on the side of a road, with its door open. The car is sleek and modern, with a black license plate that reads \"BOS\". The car is parked next to a tree, and the scene is set in a forested area. The car's door is open, suggesting that someone might be getting in or out of the car. The overall atmosphere of the video is calm and serene, with the car being the main focus. The forested area provides a natural backdrop to the scene, adding a sense of tranquility to the video. The car's sleek design and the open door suggest that it might be a high-end sports car, possibly used for leisure or racing. The video does not contain any text or additional objects, and the focus is solely on the car and its immediate surroundings."
+    ],
+    "video_ids": [
+        "O4WJvXO2HDQ_70_0to104"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white sports car with a black license plate reading 'BOS', an open door, and a tree? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white sports car with a black license plate reading 'BOS', an open door, and a tree in the foreground. The car is clearly visible, and the description matches the visual content without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a white sports car with a black license plate that 'BOS', an open door, and a tree. the background. The car is positioned white, and the license matches the core elements of any.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a18d6810621c1b51486d1cb08d331acc5a2c2b64
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac541cd4b8e37e8393e4b9016cda753f229a493e5e0576113f0c563948cec5eb
+size 156870
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..0d2c250b02de9192c0a7a7335f00c29323fa46e4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a car show presentation featuring a man in a black polo shirt with a red logo on the left chest. He is standing in front of a white sports car with a black roof and a sleek design. The man is gesturing with his right hand, possibly explaining the features of the car. The background shows a large crowd of people, indicating that the event is well-attended. The lighting is bright, with spotlights illuminating the car and the presenter. The style of the video is informative and promotional, aimed at showcasing the car and its features to potential buyers or enthusiasts."
+    ],
+    "video_ids": [
+        "VW-NXFTWPA8_4_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black polo shirt, a white sports car with a black roof, a large crowd of people.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a black polo shirt standing in front of a white sports car with a black roof. In the background, there is a large crowd of people, which matches the description. The scene appears to be at an auto show or similar event, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a black polo shirt standing in front of a white sports car with a black roof. There the background, there is a large crowd of people, which ful the description. The presence appears to be set a event show or similar event, as the elements elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f3eaea082b23bf652b1d4c25d785e8344f5a5283
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7ee1f92f72ee3f4fd41293fb33b8680db07dce29b877532af005f638e1a4e9a1
+size 178671
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..23d0541816ec3076babdb85f0fbff356fc87c1d7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features two men sitting in front of a backdrop with the title \"Avengers: Endgame\" displayed. The man on the left is wearing a blue shirt and glasses, while the man on the right is dressed in a black jacket and a red shirt. They appear to be engaged in a conversation, with the man on the right gesturing with his hand. The overall style of the video suggests it could be a promotional interview or discussion related to the movie \"Avengers: Endgame\"."
+    ],
+    "video_ids": [
+        "itKE0FulmBA_45_16to186"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men sitting and conversing in front of a backdrop with the 'Avengers: Endgame' logo. Both individuals are visible and occupy the primary focus of the frame, fulfilling the 'Two men' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two men sitting side engaginging. front of a backdrop. the textAVien' Agegame' logo. The men are visible and appear the central focus of the video, fulfilling the 'Object men' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..368701042cc55c2ce64ecdcdd5c043208c55ee3f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9760ac8f42f9aa749b3283bce68fb7bb7ddb0f176e94fb1af1043b76c556db33
+size 185927
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..238f3db245ae519d7693d66bdb487552fac85fcc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a robotic figure with a human-like form, standing upright and facing the camera. The robot is predominantly white and appears to be made of a lightweight material, possibly plastic or fiberglass. It has a head, torso, arms, and legs, with visible joints and articulation points. The robot's body is adorned with various stickers and labels, suggesting it may be a prototype or a model on display. The background is a simple, nondescript room with a tiled floor and a white wall, which provides a neutral backdrop that highlights the robot. The style of the video is straightforward and documentary, focusing on the robot without any additional context or embellishment. The camera angle is slightly elevated, looking down at the robot, which gives a sense of scale and perspective. The lighting is even, with no harsh shadows or highlights, which allows for a clear view of the robot's features and details."
+    ],
+    "video_ids": [
+        "5d4KpqZ75gk_106_0to110"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A robotic figure with a human-like form, made of a lightweight material like plastic or fiberglass.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a robotic figure with a human-like form, featuring articulated limbs and a head with facial features. The visible construction includes exposed wiring and mechanical components, suggesting lightweight materials like plastic or fiberglass are used in its build. The robot's appearance and structure align with the description, even though it also has visible branding and electronic components, which do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a robotic figure with a human-like form, which a joints and a humanoid, a features. The material material and a joints and mechanical components, which a materials like plastic or fiberglass. used in its construction. The figure's design and structure align with the description of fulfilling though the is includes a electronic and a elements, which are not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0a98782eeac00efdea0790fc22d5500b10865e27
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b14853c3110dd6bcaa8fb6ee91ce8bde096341793b281bc10144d04f070d52a5
+size 129283
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..a8047ce265ed773a977a3514c81ca693e08d613a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a delicious dessert, specifically a cake, being prepared. The cake is a layered dessert with a creamy white filling, topped with a rich chocolate ganache. The ganache is drizzled over the top of the cake, creating a beautiful pattern. The cake is also adorned with fresh raspberries, adding a pop of color and a hint of tartness to the sweet dessert. The raspberries are scattered across the top of the cake, providing a contrast to the smooth texture of the ganache. The video captures the process of the cake being prepared, from the initial drizzling of the ganache to the final placement of the raspberries. The focus is on the cake, with no other objects in the frame. The style of the video is simple and straightforward, focusing on the dessert and the process of its preparation."
+    ],
+    "video_ids": [
+        "aZDBx6aDhYA_0_0to151"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Cake, creamy white filling, rich chocolate ganache, fresh raspberries? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a cake with a creamy white filling, rich chocolate ganache drizzled over the top, and fresh raspberries as a topping. These elements are clearly visible and match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a close with creamy creamy white filling, rich chocolate ganache,izzled over it top, and fresh raspberries as a garn. The elements match consistent visible and match the description provided.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..64f9e6443ef6d4d53a15a93793d88a43fdafe777
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1145-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fdc2c4f549c59f6dfef133c09d6e8a99c5281cc25af0c963474f772d6651ea01
+size 198457
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..e18ef5eeb2040bfc460499bc59d901dee94e4ccd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man stands confidently in a field, surrounded by a group of horses. He is dressed in a blue jacket and a blue shirt, and his hands are clasped together in front of him. The horses, varying in color and size, are scattered around him, some closer and others further away. The field they are in is lush and green, with a few trees scattered around the edges. In the background, a mountain looms, adding a sense of depth and scale to the scene. The man's position in the center of the frame, along with his confident stance, suggests that he is the main subject of the video. The horses, with their varied colors and sizes, add a dynamic element to the scene. The mountain in the background provides a sense of location and context, suggesting that this scene is taking place in a rural or wilderness setting. Overall, the video captures a moment of calm and connection between the man and the horses, set against the backdrop of a beautiful natural landscape."
+    ],
+    "video_ids": [
+        "PCBl2Ec-Ww0_15_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a group of horses.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man standing in the foreground and a group of horses in the background, which fulfills the 'Object(s)' condition. The setting is a mountainous area, and the man is gesturing while speaking, which is consistent with the description. There are no elements that contradict the presence of a man and a group of horses.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a man standing in a foreground with a group of horses in the background. fulfilling alignfills the 'Object(s)' condition. The presence includes a naturalous landscape with and the presence is dresseduring, standing, which adds a with the description of The are no additional in contradict the core of a man and a group of horses.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..095092d1dc9bca58604d2647715eac198f1ed09c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f03217a144899266fc08a5a9c6f514ee938e85deeed4a14575936878f810d22
+size 183926
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..ceefa9947481b083c24e275d20a2c6dcd2494ea0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a plaid shirt, who appears to be in a room with a white wall and a door. The man is looking directly at the camera, and his facial expression seems to be one of surprise or shock. The lighting in the room is bright, and the man's face is well-lit. The style of the video is a close-up shot, focusing on the man's face and upper body. The overall mood of the video is intense and dramatic, as the man's expression conveys a strong emotion."
+    ],
+    "video_ids": [
+        "KwbQSMg0Cig_4_0to154"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a plaid shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a plaid shirt, which matches the 'Object(s)' condition. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing a plaid shirt, which matches the descriptionObject(s)' condition. The shirt and other elements in not contradict this description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..27f65d7de8653bbbcc572752834df1b5e8bf13f9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec9dc4c3d627a7ccf35dc067645894f850054fc66bff590bba8e00363a832172
+size 130373
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..d4aeda67546fe7b508318ae08cc3f8302c0574d2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a scenic drive along a winding coastal road. The road, which is the main focus of the video, is surrounded by lush greenery and rugged cliffs. The ocean, visible in the background, adds a sense of tranquility to the scene. The video is shot from a high angle, providing a bird's eye view of the road and the surrounding landscape. The colors in the video are vibrant, with the green of the trees and the blue of the ocean creating a striking contrast. The overall style of the video is serene and picturesque, capturing the beauty of nature and the joy of a road trip."
+    ],
+    "video_ids": [
+        "rOTvRjAxCkw_14_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Winding coastal road, lush greenery, rugged cliffs, ocean? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a winding coastal road running alongside the ocean, with lush greenery covering the surrounding cliffs and hills. The ocean is visible with waves crashing against rocky shores, matching the described elements. Additional text overlays do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a winding coastal road that parallel lush ocean, with lush greenery and the landscape landscape. vegetation. The rugged is visible in its crashing against the cliffs, and the description elements. The elementsural or not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1b985a66b92692f1c575d8a4868fb4197c45cc64
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6bb646f403be9779c14f4dbebf3766ab91bd455fc81050caa41220ec66cb8b5
+size 293239
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..3458eed50a8418c0daca8b9be32508fa489ed8c5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a brown bear is seen sitting on a striped chair in a cozy living room. The bear is watching a small purple doll sitting on the floor, who is also watching the bear. The room is filled with warm, inviting furniture, including a couch and a television. The television is turned on, adding to the homely atmosphere. The bear and the doll seem to be engaged in a silent conversation, creating a sense of companionship and warmth. The overall style of the video is cartoonish and whimsical, with a focus on the interaction between the bear and the doll."
+    ],
+    "video_ids": [
+        "3bpmnBd6_Ys_192_90to217"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A brown bear and a small purple doll? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully features a brown bear and a small purple doll (Masha) as described. The bear appears in the scene, initially moving into frame and then resting on the armchair, while Masha, the small purple doll, is seen sitting on the chair and later playing with a ball. These elements are central to the video and do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a brown bear and a small purple doll.whichasha) in the. The bear is to a foreground, sitting sitting slightly a, then sitting in a arm of. while theasha is a purple purple doll, is seated sitting on the floor. later on with a toy. The elements match consistent to the video's align not contradict the given.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..45cb6c7396267d5a2c00feacd7a38535c3586638
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:842f5786b1c295a35b7c17ddba4576654a9a5df8deeb3a0d04f6d3b44271d97b
+size 97975
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..9bff81054c112ca28473013b5cd8d296bde79e29
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a white hatchback car parked on the side of a road. The car has gold rims and a small orange light on the side. The car is parked next to a fence and a grassy area. In the background, there are buildings and a gas station. The sky is overcast, and the overall atmosphere of the video is calm and quiet. The car appears to be in good condition and is the main focus of the video. The video does not contain any people or animals. The style of the video is straightforward and documentary-like, with no special effects or artistic filters. The video is likely meant to showcase the car and its surroundings."
+    ],
+    "video_ids": [
+        "FRhrNWJDrGg_15_0to185"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white hatchback car, gold rims, a small orange light on the side? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a white hatchback car with gold rims and a small orange light on the side, matching the description. The car is the central focus, and no conflicting elements are present that contradict these core features.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a white hatchback car with gold rims. a small orange light on the side. which the description provided The car is parked central focus, and the additional elements are present.\"\n would the features features.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d177a8f70ea113306f39c88a4de6eef4a596ab77
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3c78e577d4a247a2bb15e1918df4984e9e9d618bfc328c7082a0e5227c2a730
+size 218030
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..2340f6add10b8c27268e40ca3f2fb989351374f2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young man is seen in a bedroom setting, wearing a gray hoodie with the hood up. He is holding a black flip phone to his ear, engaged in a conversation. His expression is one of surprise or shock, as he looks off to the side, away from the camera. The room around him is filled with various objects, including a laptop on a desk in the background, a bed, and a hat hanging on the wall. The overall style of the video is casual and candid, capturing a moment of the young man's life in his personal space."
+    ],
+    "video_ids": [
+        "_bSbR2KKtEE_1_0to199"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man, a black flip phone, a laptop, a bed, and a hat.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man holding a black flip phone, with a laptop visible in the background on a desk, a bed partially visible in the foreground, and a hat worn by a figurine in the background. All specified objects are present and do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a young man wearing a black flip phone to which a laptop in in front foreground. a desk. and hat is visible in the background, and a hat on by the personine or the background. The elements elements are present, do not contradict the description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c261eafb6a23e7d84c1d96e1551067a8ab2d1ea1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a06d42decd511e48c047f27d4a95f0cfda505a5d1afbbc38ad4d185aba03c5c
+size 174941
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..4c6db44977c975767152c0ac70026e27884549c9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen enjoying a meal in a restaurant. He is wearing a black t-shirt with the words \"I travel for food\" written on it. The man is seated at a table, which is adorned with a blue tablecloth. In front of him is a plate of food, which he is about to take a bite of. The restaurant has a blue tiled wall in the background, adding to the ambiance of the setting. The man's enjoyment of the meal suggests that he is indeed a food enthusiast who travels to experience different cuisines."
+    ],
+    "video_ids": [
+        "CodtdeXZhJA_18_0to139"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a plate of food, a black t-shirt with 'I travel for food' written on it.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black t-shirt with the text 'I travel for food' clearly visible. He is eating a piece of food, which appears to be a fried item. The setting includes a plate of food on the table, and the man is the central focus of the video. Additional elements like other people in the background do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black t-shirt with the text 'I travel for food' on visible. He is seated from plate of food from which is to be a plate item, The plate includes a plate of food in a table, and the background is seated central figure of the video. The elements like the food or the background or not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b42d2ae16cda384f1d1c8da300155a9906268a56
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:550efe0264c7315094073fdaf870b5b9fe5d45514eaeaecd196a5e9031c680f8
+size 147270
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..d91f78bbec7f2de9033ff57fd0793a79458daeab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a black BMW sports car in motion on a winding road. The car is sleek and shiny, with a distinctive front grille and headlights. The road is surrounded by lush greenery and rocky cliffs, creating a picturesque backdrop for the car's journey. The car's speed is evident in the blurred background, suggesting a thrilling ride. The overall style of the video is dynamic and exciting, with a focus on the car's performance and the beauty of the natural surroundings."
+    ],
+    "video_ids": [
+        "DImA1KI4HHo_2_74to286"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black BMW sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a black BMW sports car driving on a road, matching the description. The car's design, color, and brand are consistent with a BMW sports model, and there are no conflicting elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a black BMW sports car, on a road. which the description of The car's design, including, and the are consistent with the BMW sports car. and the are no elements elements in would the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bde26bd8bf4065ee7c2e01fa50cbac21015c249f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1146-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fbee1baf5e0983d6894d6b4e9e20ef1d4b45e2450817a2210f61c59f3035dc87
+size 251925
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..0c7260f71804a2dec68a3f34ed6a87f352dd6e31
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a rugby match. The main focus is a player in a red and white jersey, who is in possession of the rugby ball. He is in the midst of a powerful run, with his arms outstretched and his gaze focused on the goal. His body language suggests he is determined to score a try.  In pursuit of the player with the ball are two defenders, both in black jerseys. They are running towards the player, their bodies leaning forward in an attempt to catch up and tackle him. Their eyes are locked on the ball, indicating their intent to prevent a score.  The background of the video is a rugby field, with the goal line visible in the distance. The field is well-maintained, with the grass cut short and the lines clearly marked. The lighting suggests it's either late afternoon or early evening, casting long shadows and creating a dramatic atmosphere.  Overall, the video is a snapshot of a high-stakes moment in a rugby match, with the outcome hanging in the balance as the player with the ball tries to evade the defenders and score a try."
+    ],
+    "video_ids": [
+        "CA_OMjgcLM8_4_23to189"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a red and white jersey with a rugby ball, two defenders in black jerseys, and the rugby field? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a red and white jersey holding a rugby ball, with two defenders in black jerseys nearby. The setting is a rugby field, matching the core description. Additional elements like the 'news24VIDEO' logo and player name overlay do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a player in a red and white jersey holding a rugby ball, with two players in black jerseys running. The setting is a rugby field, which the description description provided The elements such the refereeS'4'' watermark and the numbers are do not contradict the main and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..026f0ba3ee2918ac5246a8379407765e8edbba1e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e6f2f8de53924e6d8620dbb2c05335365f951fb5ef765cdd562a7cd1cf28e34
+size 261841
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..89499f850a93329849c072fb99806e158afe742a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a delightful culinary experience, featuring a variety of sushi dishes presented on a clear glass plate. The plate is placed on a wooden table, adding a rustic charm to the setting. The sushi includes a selection of fresh fish, including salmon and tuna, as well as a few pieces of cucumber and tomato, adding a touch of freshness to the dish. The sushi is arranged neatly on the plate, with each piece carefully placed next to the others. The colors of the sushi contrast beautifully with the clear glass plate, making the dish look even more appetizing. The overall style of the video is simple yet elegant, focusing on the food and the presentation rather than any additional elements. The video does not contain any text or other objects, keeping the viewer's attention solely on the sushi dish. The video is likely to be used for promoting a sushi restaurant or showcasing the skills of a sushi chef."
+    ],
+    "video_ids": [
+        "ggfvsoS9xS8_19_0to174"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Sushi dishes, clear glass plate, wooden table, fresh fish (salmon and tuna), cucumber, tomato? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing sushi dishes, including fresh salmon and tuna sashimi, on a clear glass plate. The plate is placed on a wooden table, and cucumber slices are visible as garnish. Although tomato is not clearly identifiable as a separate element, the red garnish next to the avocado could be interpreted as a tomato-based condiment or garnish, which aligns with the description. The overall scene matches the specified elements without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showcasing showing sushi dishes, a pieces fish and tuna,ashimi, on a clear glass plate. The plate is placed on a wooden table, and there and and visible, garnish. There there slices not explicitly visible in a separate object, it description colorish on to the sushi slices be mistaken as a tomato slice sauceiment or garnish, which iss with the description. The presence scene matches the provided objects without any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..83a44d3f8ecbc1baf63e3d98aba49128454ed9b2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8cb084c9e8d9050cbde42211cf6a867c80d3971ba0d24eaf1bad8cbc5283b4e3
+size 95195
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..dd83ba5dcc7a73fad57fb2cd6f09ca72c069ce11
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view inside an aquarium, focusing on the bubbling water surface. The bubbles are numerous and vary in size, creating a dynamic and textured appearance. In the background, a fish is visible, swimming gracefully near the top of the tank. The fish appears to be a type of tetra, characterized by its elongated body and vibrant coloration. The tank's environment includes a blue and white striped object, possibly a filter or decoration, which adds a contrasting backdrop to the scene. The camera remains stationary throughout the sequence, maintaining a consistent focus on the interaction between the bubbles and the fish. The overall atmosphere is serene, highlighting the natural behavior of the fish amidst the gentle movement of the bubbles."
+    ],
+    "video_ids": [
+        "1c0c93fdff4b2afa0259a56523965e479a7a4fa506d3f7cbf0ecabb73a3fe1f5"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bubbles, a fish (likely a tetra), and a blue and white striped object.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows bubbles rising in the water, a small fish (consistent with a tetra) swimming in the background, and a blue and white striped object (likely a decorative element or part of the tank's design). These elements align with the described 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts shows bubbles,, the water, a fish fish thatlikely with a tetra) swimming in the background, and a blue and white striped object,possibly a toy item or part of the tank setup design) The elements match with the ' 'Object(s)' condition,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..280d1f76abe53083749edcb528d64d315a435074
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b3a23a3541352a36c269921594393ceb34863bb807781cbb33eafdb81b5dce05
+size 317387
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..491d5ebc562fe98313264e861d0330c57caa3152
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated scene set in a historical or mythological setting. It features a group of characters dressed in ancient Roman or Greek attire, standing in a courtyard with classical architecture in the background. The characters are engaged in a conversation, with one character gesturing towards another. The scene is rich in detail, with realistic textures and lighting that give it a lifelike appearance. The characters are depicted with realistic facial expressions and body language, suggesting a serious or intense discussion. The overall style of the video is realistic with a focus on historical accuracy, and it is likely designed to be part of a larger narrative or story."
+    ],
+    "video_ids": [
+        "I2A0f_PUXfo_7_122to277"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A group of characters dressed in ancient Roman or Greek attire.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a group of characters dressed in attire consistent with ancient Greek or Roman styles, including tunics, cloaks, and armor. The setting, with its classical architecture and palm trees, further supports the ancient Mediterranean context. No elements contradict the description of the characters' attire.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a group of characters dressed in attire that with ancient Roman or Roman styles. including togics, cloaks, and head. The setting also with its classical architecture featuring columns trees, also supports the ancient Greek theme. The elements contradict the description of ancient '' attire or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5322e5355f69728dc9283e5566d90fb1e50f317f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb9e3ea37282b5313ecd7626d7b13a9b621484dcd91d138181d99346217d1755
+size 144756
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..26e5c1bb822c16ec6a67feead9843918785f1c57
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman with long brown hair is seen in a brightly lit room with a white wall and a red plant in the background. She is wearing a blue floral blouse and has a surprised expression on her face. Her eyes are wide open and her mouth is slightly open as if she is in the middle of a conversation or reacting to something. The overall style of the video is casual and candid, capturing a moment of surprise or interest in the woman's expression."
+    ],
+    "video_ids": [
+        "6BO2xOjooKY_4_20to197"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with long brown hair, wearing a blue floral blouse.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with long brown hair wearing a blue floral blouse, which matches the description. The background elements, such as the plant and wall decor, do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with long brown hair wearing a blue floral blouse. which matches the description provided The background includes, such as the plant, the,, do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..33fc825f1cc4935ab4cbf15fd6ad83f75fa941f2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e18f0b7461ed939e8745c02742fd40d5a22d8298e4aa6b337a419ee36e00b0e3
+size 165892
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..ba8350e0596c957dd382e35966c7d4f926059ae7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are seated at a dining table, enjoying a meal together. The man is wearing a red bandana on his head and a blue shirt, while the woman is dressed in a white blouse. They are both using chopsticks to eat from their bowls, which contain a noodle dish. The table is set with various condiments, including a yellow basket filled with lettuce and a jar of sauce. The setting appears to be a casual dining environment, possibly a restaurant or a home kitchen. The atmosphere is relaxed and intimate, with the focus on the shared meal and conversation between the two individuals."
+    ],
+    "video_ids": [
+        "T0gwaAP5k9Y_87_47to189"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman sitting at a table, eating noodles with chopsticks and spoons. Both individuals are visible and central to the scene, fulfilling the 'Object(s): A man and a woman' condition. Additional elements like food, jars, and background decor do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two man and a woman sitting at a table, eating from with chopsticks. smilingons. The individuals are wearing, appear to the scene, fulfilling the 'Object(s)' A man and a woman' condition. The elements like the and utens, and a details do not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..864c529f807b99408dcc7113c801b360685738a2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1c24a675a3f3ae4ffbeaf4d4a1f18a66184013de81ca20afae0f6e1db771ed6d
+size 185773
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..f8237a54ba0242d3601810d3eb9c8332b8baf3d0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a blonde doll with long hair, wearing a blue top and purple pants, standing in a kitchen. She is interacting with a blender on a white countertop. The kitchen has red cabinets and a white refrigerator. The doll is seen in three different actions: first, she is pouring a red liquid from a bottle into the blender; second, she is holding a red spoon and stirring the contents of the blender; and third, she is holding a red spatula and appears to be serving the contents of the blender into a bowl. The doll's actions suggest that she is preparing a meal or a drink. The style of the video is realistic, with attention to detail in the doll's movements and the kitchen's setting."
+    ],
+    "video_ids": [
+        "SgT7pGPONTw_32_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blonde doll with long hair, wearing a blue top and purple pants. A blender, a red liquid bottle, a red spoon, a red spatula, and a bowl.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a blonde doll with long hair, wearing a blue top and purple pants, standing in a kitchen. A blender is visible on the counter, and there is a bowl with fruit next to it. The description mentions a red liquid bottle, red spoon, and red spatula, but these items are not visible in the video. Since the core elements (doll, blender, bowl) are present and there are no contradictions, the video largely matches the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a blonde doll with long hair, wearing a blue top and purple pants, which in a kitchen. The blender is present on the counter, and a is a red and a being to it. A doll mentions a red liquid bottle, a spoon, and red spatula, which these are are not clearly in the video. The the core elements matchthe, blender, and, match present and match are no contradictions, the video largely ful the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9d20000ee33f7be3016bc1fba8dab17720cafcac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:acd084ee0d923c33d5adb5ed674b10d00142c595b04856ec530b9fdfe18272c0
+size 102682
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..1bd7541a19245f7d7035fb824035e6119158a21a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a vibrant blue betta fish swimming gracefully inside a clear wine glass filled with water. The fish's flowing fins and tail move fluidly as it navigates the confined space. The background is a plain white surface, which contrasts sharply with the deep blue color of the fish, making it stand out prominently. The lighting is bright, highlighting the fish's iridescent scales and the clarity of the water. The fish appears to be exploring its surroundings, occasionally pausing to adjust its position or swim in different directions. The overall scene is serene and visually striking, focusing entirely on the graceful movements of the betta fish within the glass. There are no significant changes in the environment or the actions of the fish throughout the video; it remains a consistent and tranquil depiction of the fish in its aquatic habitat."
+    ],
+    "video_ids": [
+        "6e2a6ba0e078410d5c7d2a0e2e413dd66292586fae1920c2305cdc26ce08ecba"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A vibrant blue betta fish? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a vibrant blue betta fish swimming inside a wine glass filled with water, which matches the core description. The fish's color and species are accurately represented, and no elements contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a vibrant blue betta fish swimming in a clear glass, with water. which align the description description of The fish's color and the are clearly represented, and the additional contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1e0cfbbba9091edef38edd764f5d6a888fb0d24f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1147-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ccd72e797b4d23daa4381d6df961e94aa6e2124a5cb593fd1fbbe8ad1795c274
+size 206914
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..338cdeaa8fac4ab2e7e8a8af455f23e590503c06
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a black sports car parked in a garage. The car is sleek and shiny, with a large wheel and a distinctive design. The garage is spacious and well-lit, with a high ceiling and a concrete floor. The car is the main focus of the video, and it is shown from different angles, highlighting its design and features. The video is likely a promotional or sales video, showcasing the car's design and features to potential buyers. The style of the video is professional and polished, with high-quality footage and a clear focus on the car."
+    ],
+    "video_ids": [
+        "wVS3hoqBVdg_23_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a black sports car, specifically a Mercedes-AMG SLS, which matches the description. The car is clearly visible and occupies the central focus of the frames, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a black sports car, which focusing close-BenzAMG GT6, which is the description of The car is shown visible, the the majority focus of the frames. with the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6d633f3251c5336ad78c89b49c997747affe3f8c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a6943baeed8aea74632ad2660bb5b4780e0eae9039bb2ea84141f180f51d8e93
+size 186993
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..3f22c3e0afc15c5b54cd5fff69f9bd2a29f746f7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and glasses, wearing a blue shirt. He is standing against a red background. The man appears to be speaking or presenting, as he is looking directly at the camera. The style of the video is simple and straightforward, with no additional elements or distractions. The focus is solely on the man and his message. The lighting is even, and the colors are vibrant, making the man stand out against the red background. The overall impression is that of a professional or educational video."
+    ],
+    "video_ids": [
+        "VWuykQAt82Y_1_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man with a beard and glasses, wearing a blue shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man with a beard and glasses, wearing a blue shirt, which matches the description. There are no conflicting elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man with a beard and glasses, wearing a blue shirt. which matches the description provided The are no additional elements in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e56eb57024b05cf00364aa34349e5c3f7e6cb55a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:beb5f52a07a50844d711b6b66ef1a379e856969aa807ca9bec7d7dfad37dcc7d
+size 195513
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..9b5ac229a6bc48d201c9a7a6964a359197d9f750
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a suit standing inside an airplane cabin. The man is smiling and appears to be speaking. The airplane cabin is well-lit, with natural light coming through the windows. The windows are round and provide a view of the sky outside. The man is standing in front of a row of seats, which are empty. The style of the video is candid and informal, capturing a moment in the man's journey."
+    ],
+    "video_ids": [
+        "OdzQ0Tb5DqY_3_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit, a row of empty seats? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man dressed in a suit standing inside what appears to be a private jet. Behind him, there is a row of empty, white leather seats, which matches the described elements. The setting and objects are consistent with the 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man in in a suit, in an appears to be an private jet cabin There him, there is a row of empty seats white leather seats, which align the description '. The setting and the in consistent with the givenObject(s)' condition provided any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7dd1dd8a5c4caea451e8c53fed14c4c1fadcd6c9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd523c8d2339b5ce183f80ef5bbc481d9dec2589dafef98b6f254265647a6420
+size 101584
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..b4c71de034bc6627dced26b7a445b64e3100c5c2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young woman is being interviewed by a news reporter. She is wearing a blue jacket and a white headband. The interview is taking place in a park with trees in the background. The woman is holding a microphone and appears to be speaking. The reporter is holding a camera and is also wearing a microphone. The video is shot in a realistic style, capturing the interaction between the woman and the reporter. The park setting adds a natural element to the scene."
+    ],
+    "video_ids": [
+        "4WBNGTt6j8w_5_81to209"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman, a news reporter, a microphone, a camera.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman being interviewed outdoors, holding a microphone with the 'NSF' logo, which implies she is being interviewed by a news reporter. The presence of a camera operator (partially visible on the right) further supports the presence of a news reporter and camera. All core elements described are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a young woman wearing interviewed,, which a microphone and a logoTN' logo, which suggests she is a treated by a news reporter. The presence of a camera in holdingpartially visible) the left) and supports the condition of a news reporter conducting the, The elements elements of in present and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..81c70057aa2c17193abc5c7b38df3759cf1231d8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:890006d7c073c62a5ad63e5bf9a12463e95144f6d98f50866506d532b001bfaa
+size 141513
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..5fe2b27de3a02b0e79694f10c8f6a2ce1cab3a06
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a white Mercedes-Benz SUV on display at a car show. The vehicle is positioned on a stage with a black floor and a curved backdrop. The SUV is equipped with a black grille, black rims, and a black roof rack. The car is presented in three different angles, highlighting its design and features. The lighting in the video is dim, with the car being the main focus. The overall style of the video is sleek and professional, emphasizing the luxury and sophistication of the Mercedes-Benz SUV."
+    ],
+    "video_ids": [
+        "p8AkZXtyxcI_0_0to104"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white Mercedes-Benz SUV? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a white Mercedes-Benz SUV, which is clearly visible and central to the scene. The vehicle's design, including the Mercedes-Benz emblem and distinctive grille, matches the description. Additional elements like another car and background signage do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a white Mercedes-Benz SUV, which matches the visible in matches to the frames. The vehicle's design, including its box-Benz logo and the SUV, is the description of The elements like the vehicle in a details do not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bbbe397c984d5d521f474b30bfffb377fcecfca8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:beb272cb3860d92eacacc5863c2f4f852b66b8c2d6e1286784945bbd874ed0d6
+size 63210
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..bbd3793005a98c774ecb1b33a125191218bda230
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the intricate details of a motorcycle engine, showcasing its mechanical components and design. The engine is the focal point of the video, with its shiny metal parts and complex structure. The motorcycle is parked in a showroom, with other motorcycles visible in the background. The lighting in the showroom highlights the motorcycle's details, creating a sense of depth and dimension. The video is a close-up shot, focusing on the engine and its components, providing a detailed view of the motorcycle's inner workings. The style of the video is informative and technical, providing viewers with a comprehensive look at the motorcycle's engine."
+    ],
+    "video_ids": [
+        "3YP-7EVUZ5k_8_0to122"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A motorcycle engine? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a close-up of a motorcycle engine, showcasing its components such as the cylinders, bolts, and surrounding parts. The engine is clearly visible and is the central focus of the video, fulfilling the 'Object(s)' condition of a motorcycle engine.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a detailed-up view a motorcycle engine, which its components such as the cylinders, air, and other parts. The detailed is the the and occupies the central focus of the video, fulfilling the 'Object(s)' condition of depicting motorcycle engine.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c70ce5efee41b13d4291017fe8d73657e40649ef
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:857deb2f87ceacc06353d0b05ab71255549cb22f9034c1494106294703bda451
+size 170686
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..5173fc20d29f9ed4951b7e26f59f89dbb12f4be2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman with blonde hair is seen in a close-up shot. She is wearing a black top and gold hoop earrings. The woman appears to be in a room with a red wall in the background. The lighting in the room is dim, creating a moody atmosphere. The woman's expression is serious, suggesting that she is engaged in an important conversation or situation. The overall style of the video is realistic, with a focus on the woman's face and the room's interior."
+    ],
+    "video_ids": [
+        "6Gcj48Njch4_10_69to232"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with blonde hair wearing a black top and gold hoop earrings.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair, wearing a black top and gold hoop earrings, which matches the description. The background and lighting are consistent with an indoor setting, and no elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair, wearing a black top, gold hoop earrings. which matches the description provided The background and additional are consistent with the indoor setting, and there additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e734093385cbc5cdc5698bc7a415798641f618b7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1081a31781021f17717d5ec66c7de5d439362be5f7a50687acef71d201da0d80
+size 125631
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..a946c5fdce1e1fd315c9a5da7713b8b8f69aa1a5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a dynamic and colorful scene set in a parking lot filled with various cars. The main focus is a man in a black t-shirt and a baseball cap, who is walking towards the camera with a surprised expression on his face. He is wearing sunglasses and has a beard. The man is walking past a silver sports car with the top down, which is parked next to a black sports car. The cars are sleek and shiny, reflecting the bright sunlight. In the background, there are other cars parked, including a white sedan and a black SUV. The parking lot is bustling with activity, and there are people walking around. The sky is clear and blue, suggesting a sunny day. The overall style of the video is energetic and vibrant, capturing the excitement of the man and the beauty of the cars."
+    ],
+    "video_ids": [
+        "Jl1ZF3C4Oyo_31_0to131"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black t-shirt and baseball cap, a silver sports car with the top down, a black sports car, a white sedan, a black SUV, and other people walking around.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black t-shirt and baseball cap, standing in front of a silver sports car with the top down (visible at 0:00-0:09). A black sports car is also visible to the right (0:00-0:09). A white sedan is parked in the background (0:00-0:09), and a black SUV is visible further back (0:00-0:09). Several people are seen walking around in the background (0:00-0:09). All core elements from the description are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts a man wearing a black t-shirt and a cap, walking in a of a silver sports car with the top down,which in the0:00-0:05). There black sports car, also visible in the right of0:00-0:09). A white sedan and parked in the background (0:00-0:09). and a black SUV is parked further back (0:00-0:09). There other are walking walking around in the background,0:00-0:09). The these elements of the description are present in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c8f56638e89c25a155ea8110da6dff96a72ce1ed
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1148-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fbe6bac9b26da67c277286132a7e611a5f95d295668d144835a1b7440141c9d7
+size 249569
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..cba3861657c48b4425f256e4adc520f49bbc5d95
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is on two players, one in a maroon jersey with the number 4 and the other in a white jersey with the number 51. The player in the maroon jersey is in the process of throwing the football, his arm extended and the ball just released from his hand. The player in the white jersey is in pursuit, running towards the quarterback with his arms outstretched, ready to make a tackle. The background is filled with the blurred figures of other players and spectators, adding to the sense of action and excitement. The video is shot from a side angle, providing a clear view of the players and their actions. The style of the video is realistic, capturing the intensity and speed of the game."
+    ],
+    "video_ids": [
+        "vwAUa67Ix3Y_6_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two players - one in a maroon jersey with number 4 and the other in a white jersey with number 51.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two players: one in a maroon jersey with the number 4 and another in a white jersey with the number 51. Both players are visible and their jerseys match the description. The presence of additional elements like the text overlay and background does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two players, one in a maroon jersey with the number 4 and another in a white jersey with the number 51. The players are actively and engaged jerseys match the description. The video of other players like other football on and the figures not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..371d98aee8695187135947e312c407cb868312c5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:836bae7e68c892542d63270d3cfeeaa5aaba4c84ea7360747bf64c3525776054
+size 265339
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..883d46c82511ae4a50070b73951e4d491ee971ce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a vibrant green Porsche sports car parked in front of a building with a large window. The car is positioned at an angle, showcasing its sleek design and black rims. The building's window reflects the car, adding depth to the scene. The car's shiny exterior reflects the surrounding environment, indicating that the video was taken during the day. The overall style of the video is dynamic and colorful, emphasizing the car's design and the urban setting."
+    ],
+    "video_ids": [
+        "mOXXjUpaFd8_61_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A vibrant green Porsche sports car, a building with a large window? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a vibrant green Porsche sports car, specifically a Porsche 911 GT3, parked outside a building with a large window. The car is clearly visible, and the building's facade with its large glass window is also evident in the background, matching the description. Additional elements, such as a person in a red hoodie and a poster of a race car, do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a vibrant green Porsche sports car, which a model Cay711,3, which in a building with a large window. The car's the visible, and the building's large, the large window window is also evident in the background. fulfilling the description provided The elements like such as the brick walking the car shirt and a parked on a car car, do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3fa6d0565ad219e17811a7177227447d41132587
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d2738161e168ec951a1aeb06c4e434acee43ef7dd28d8b8e62457fb0e89c44d7
+size 145129
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..5760788106d9e4f4c7e149869c0fff3f19fd32e4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a basketball player in action, wearing a white and orange jersey with the word \"SUNS\" emblazoned across the front. The player is seen in three different frames, each showcasing a different moment of the game. In the first frame, the player is seen looking up, possibly tracking the ball's trajectory. In the second frame, the player is seen in a defensive stance, ready to react to the ball's movement. In the third frame, the player is seen looking down, possibly strategizing his next move. The player's jersey is clearly visible in all three frames, indicating his team affiliation. The style of the video is dynamic and action-packed, capturing the intensity and excitement of a basketball game."
+    ],
+    "video_ids": [
+        "NgRAr-1A7l0_13_0to128"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: basketball player? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a basketball player wearing a Phoenix Suns jersey, standing on a court with a focused expression. The player's attire, posture, and setting are consistent with a basketball game or practice, fulfilling the 'Object(s): basketball player' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a basketball player in a Suns Suns jersey, which on a basketball, a basketball expression, The player's attire and posture, and the are consistent with the basketball player, practice, fulfilling the 'Object(s)' basketball player' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6681eb3cd1b5eb02a73dd89132025549df829053
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86f792c472f04b262d729151b9035d910fd52518952032f7ae238f46ca1e970b
+size 316911
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..2a85c9685b1c34f82eb32488f13aa7cdd5dcb587
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a black shirt sitting in a chair with a blue background. He is gesturing with his hands, indicating that he is in the middle of a conversation or explanation. The man appears to be speaking with confidence and authority, suggesting that he is knowledgeable about the topic at hand. The overall style of the video is professional and polished, with a focus on the man and his message. The blue background provides a calm and neutral setting, allowing the viewer to focus on the man and his words."
+    ],
+    "video_ids": [
+        "50_5IsYsjpc_44_97to242"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a black shirt sitting in a chair? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black shirt, seated in a chair, which matches the core description. The background and other elements do not contradict this depiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black shirt, seated in a chair, which align the description description provided The background is the elements in not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..63c2386365db47d54af0fcefe1ae0e9d5ecfd57f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a94bb519928b52d486902003bbe9fb322a3c0984579ab2a072c90339d5412205
+size 165610
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..474d64204df6eed3008d65f46180266c94f0063d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the breathtaking view of a mountainous landscape, showcasing the grandeur of nature. The camera, positioned high above, provides a bird's eye view of the scene, allowing us to appreciate the vastness of the terrain. The mountains, with their rocky peaks and lush green vegetation, dominate the frame, their majestic presence undeniable. The perspective of the video, combined with the natural beauty of the landscape, creates a sense of awe and wonder. The image is a testament to the power and beauty of nature, captured in a single, stunning frame."
+    ],
+    "video_ids": [
+        "FOlnSV0QOe8_116_160to338"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Mountains, rocky peaks, lush green vegetation? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features towering mountains with distinct rocky peaks, covered in lush green vegetation. The visual elements align perfectly with the described conditions, showcasing a natural landscape dominated by these features without any conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features mountains mountains with rocky rocky peaks, which in patches green vegetation. The presence elements such well with the ' ', showcasing a serene landscape that by these features. any conflicting elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2e3d7a1bb1e072814b46340d6eb202bc2c455c13
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d8aee3034e87418088c0bdaff5a70edbc5099569020b4bd0bd7340693b020d86
+size 135966
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..dcf3e6e3ad51d05e6b9141551f05e8da9798a4af
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit and tie, sitting in front of an orange background. He is wearing a gray suit and a striped tie. The man appears to be speaking into a microphone, suggesting that he might be giving a speech or an interview. The style of the video is formal and professional, with the man dressed in business attire and the orange background providing a stark contrast to his outfit. The focus is on the man and his speech, with no other objects or people visible in the video. The video is likely intended for a professional or corporate audience."
+    ],
+    "video_ids": [
+        "L_UAeI9NBGI_7_16to164"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a gray suit and striped tie, speaking into a microphone? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a gray suit and a striped tie, and there is a microphone clipped to his suit. Although the man is not visibly speaking, the presence of the microphone and his attire match the description. The video does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a man in a gray suit and a striped tie, which he is a microphone visible to his suit, The the man is not speaking speaking, the presence of the microphone suggests his formal align the description provided The background does not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dfc54c813e2ce8ad35e5fcceeb1f37b2b313dba7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d6409e684f1c75c0f212b7e031ab1df705de38f3cec017ec75b6e775039a58d5
+size 122515
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..0787e70b24cd4d2db36fc759a9ae6cbcc6dabd40
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a collection of handmade gift tags and bows, arranged on a wooden surface. The tags are designed in a variety of colors and patterns, with the phrase \"FOR YOU\" prominently displayed in white text. The bows are made from purple paper and are tied around the tags, adding a decorative touch. The tags and bows are arranged in a visually appealing manner, with some tags overlapping and others placed in the background. The overall style of the video is simple and elegant, focusing on the craftsmanship of the gift tags and bows. The wooden surface provides a natural and warm backdrop for the colorful tags and bows."
+    ],
+    "video_ids": [
+        "V9Wwi8VbKIE_0_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Handmade gift tags and bows? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features handmade gift tags shaped like jars with 'FOR YOU' text and decorative bows in various colors (purple, yellow, orange), all attached to small gift boxes. These elements clearly fulfill the 'Handmade gift tags and bows' condition, even though there are additional decorative elements like flowers and leaves that do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features colorful gift tags and like stars, colorfulYOU YOU' written and colorful bows made vibrant colors,purple, green, green, which of to the gift boxes. These elements clearly fulfill the 'Objectmade gift tags and bows' condition as as though the are additional decorative elements like the and a in are not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c00c16a092b02d20484bf2c86b2830399104e8c7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4c3a7f5d1ee7da87fd61945efbd909566a4cbbdc5f77f0e12d9adb5f4735ca6
+size 211724
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..61eb50a3f9654dfc5bf7007fcfc92937918288e0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a woman in a living room, wearing glasses and a blue shirt. She is holding her hand to her neck, and her expression is one of concern or worry. The room has a couch and a window in the background. The lighting is dim, and the overall atmosphere is subdued."
+    ],
+    "video_ids": [
+        "HKMgO0jfGxg_44_0to138"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a living room, wearing glasses and a blue shirt. She is holding her hand to her neck with an expression of concern or worry.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a living room, wearing glasses and a blue shirt, holding her hand to her neck with an expression that conveys concern or worry. The background elements, such as furniture and windows, are consistent with a living room setting, and there are no contradictions with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a woman in a living room setting wearing glasses and a blue shirt. holding her hand to her neck with an expression of appearsveys concern or worry. The setting appears, such as the and a, are consistent with a living room setting. and the are no additional to the given description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f3c22912fbaa9af13a3988a3b81221a71ace86bc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1149-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9798842e2d9ef64565bb262b45f86a9dc18018e4b3cbd74ce22c749c9092c869
+size 218998
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..169b1d4e7043fa65a173a1bbf82b9558634d1c8a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a BMW car engine compartment with a focus on the engine cover. The engine cover is black with the BMW logo prominently displayed in the center. The engine cover is secured with a silver latch. The engine compartment is well-organized with various components neatly arranged. The style of the video is a close-up, static shot with a focus on the engine cover. The video does not show any movement or action. The lighting in the video is bright, illuminating the engine compartment and the engine cover. The video does not contain any text or additional graphics. The video is likely intended for educational or promotional purposes, showcasing the engine compartment of a BMW car."
+    ],
+    "video_ids": [
+        "weJs8CsPL_8_13_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: BMW car engine cover, BMW logo, silver latch? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the BMW car engine cover with the BMW logo prominently displayed on it. The silver latch mechanism is also visible, securing the engine cover. These elements are central to the video's content and are accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a engine car engine cover, the BMW logo prominently displayed in it. Additionally silver latch is is also visible, indicating the engine cover. The elements match consistent to the description's content and align consistent represented, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fb96392644b9a776487bf38640b90da0cf116fb2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7279464bbe8f0f31483d7b531ea28067273bef001849b1ae0a2a0e43787c51af
+size 195443
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..97824014ad2fed8ae206af29fedc9288126ba7ee
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a black and brown dog with a black nose and brown eyes. The dog is looking directly at the camera with a calm and attentive expression. The background is filled with red flowers, creating a vibrant and colorful contrast with the dog's dark fur. The dog's ears are perked up, indicating alertness. The overall style of the video is a close-up shot with a shallow depth of field, focusing on the dog's face while the background is blurred. The lighting is soft and natural, suggesting an outdoor setting. The video captures the dog's gentle and curious nature, making it a heartwarming and visually appealing scene."
+    ],
+    "video_ids": [
+        "SbYb83SQ2uQ_18_0to150"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Black and brown dog with a black nose and brown eyes? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a dog that is predominantly black with brown markings around the eyes, muzzle, and ears, and it has a black nose. The eyes appear to be brown, matching the description. The background of red flowers does not contradict the core description of the dog's appearance.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a close-up of a dog with has predominantly black with brown markings, its eyes and which, and chest. which it has a black nose and The dog are to be brown, which the description. The background is the flowers is not contradict the core description of the dog's appearance.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f320795d6c42ff565079936edce46137d53b83dc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc2d96c9bec50ce42eb8a351112f8c48b2ad4b017f27f1717e7297b25863f552
+size 100728
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..b697db702d385b603b90ed991d02c6b550b694b7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features an older woman with curly hair, wearing a blue shirt, sitting in a chair. She appears to be engaged in a conversation or interview. The setting is an office or a room with a bookshelf in the background. The woman is the main subject of the video, and her facial expressions and body language suggest she is actively participating in the discussion. The style of the video is a straightforward interview or conversation, with the focus on the woman's responses and the context of the setting."
+    ],
+    "video_ids": [
+        "VnEA39R3n74_38_63to197"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An older woman with curly hair, wearing a blue shirt, sitting in a chair.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an older woman with curly hair, wearing a blue shirt, and sitting in a chair, which matches the core description. Additional elements like a framed photo and bookshelf in the background do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an older woman with curly hair, wearing a blue shirt, sitting sitting in a chair. which align the description description provided The elements such the background picture or a on in the background do not contradict the main and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f2327c327bac3c3386909e3a7f1ecc95d4183e9a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b50993e5405423c2b435adfb721e4a3c63f8f000871d2e788c67a44f1076a298
+size 198318
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..1047415a521a6d93085586accce60bcf2042749c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men in suits, each raising their hands in a gesture of agreement or approval. The first man, on the left, is slightly ahead of the second man, who is on the right. Both men are wearing glasses and blue ties. The background is blurred, but it appears to be an indoor setting with a wooden paneled wall. The style of the video is a straightforward, unembellished news report, likely from a financial or business news source, as indicated by the \"Bloomberg\" logo in the bottom left corner. The focus is on the men and their actions, with no additional context or embellishments provided."
+    ],
+    "video_ids": [
+        "LLBRVZk9Ynk_4_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men in suits, both wearing glasses and blue ties.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men in suits, both wearing glasses and blue ties, as described. The scene appears to be a formal setting, possibly a hearing or deposition, where both individuals are raising their hands, consistent with the description. No elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men in suits, both wearing glasses and blue ties, which described. They men is to be a still setting, possibly a meeting or a, where individuals individuals are raising their hands, which with the ' of The additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..963968c73c9cc9261b5b580efcadd223e9a0a95b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:91318a8a40af3bbd618f157a27f7220fd3701b5c31175b5edd69a00eded6cb3b
+size 181171
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..2ae0f0efbc04765a9c7606c118d4cfce214c1745
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen preparing a meal in a kitchen. She is wearing a blue and white patterned shirt and is holding a bowl in her hand. The kitchen is well-equipped with various utensils and appliances, including a sink, a stove, and a cutting board. The woman is standing in front of a window, which allows natural light to fill the room. The overall style of the video is casual and homey, capturing the everyday activity of cooking."
+    ],
+    "video_ids": [
+        "Aon9FUiHw2U_4_20to230"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, wearing a blue and white patterned shirt, holding a bowl.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The woman in the video is clearly wearing a blue and white patterned shirt and is holding a bowl, which matches the description. The presence of other kitchen elements does not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video in the video is wearing wearing a blue and white patterned shirt. is holding a bowl. which align the description provided The setting of other kitchen items like not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6359ab2a4b3c20fb12ad9bd07fa094df397c906d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c8e87bf1489c2a4852ec33bacb0e45d675c1440be6ddb3fa1a5ad0778766b8bc
+size 221081
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..e5bcbee75bcae108b0ee21351938f2f26c082e40
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a white Jeep Wrangler driving through a grassy field. The Jeep is equipped with large off-road tires and a black front bumper. The vehicle is covered in mud, indicating that it has been used for off-road activities. The Jeep's headlights are on, illuminating the path ahead. The background features a grassy hill with trees, suggesting a rural or wilderness setting. The style of the video is realistic, with a focus on the vehicle and its surroundings. The camera angle is from the front of the Jeep, giving a sense of the driver's perspective. The video captures the Jeep's movement through the field, highlighting its off-road capabilities."
+    ],
+    "video_ids": [
+        "8Hopt4m-bwA_52_117to243"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white Jeep Wrangler? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a white Jeep Wrangler with large off-road tires, positioned on a grassy hill. The vehicle's iconic grille, round headlights, and rugged design are clearly visible, matching the description of a white Jeep Wrangler. The surrounding environment does not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a white vehicle Wrangler, a off-road tires and a in a grassy terrainside The vehicle's design design and headlights headlights, and rugged design are clearly visible, align the description of a white Jeep Wrangler. The setting environment, not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cfd7bcbdd019d9425d8ece9cd5996f2b9a2d197e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77423472dce1d8a55242eb6f57d78f01b38f36dc7e7e50d2399b07016a83d6a8
+size 213712
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..61eb8801b24e17abe652dfbb1d42eec7fded1453
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a dynamic and stylish showcase of a red sports car in motion. The car, with its sleek design and shiny exterior, is the main focus of the video. It is captured in three different frames, each highlighting a different aspect of the car's design and movement.  In the first frame, the car is seen from a side angle, showcasing its aerodynamic shape and the smooth lines that run along its body. The car's red color stands out against the backdrop of a building, drawing attention to its design.  The second frame captures the car from a front angle, emphasizing its powerful stance and the sharp angles of its headlights. The car's black rims add a touch of contrast to its red exterior.  The third frame is a close-up of the car's front, highlighting the intricate details of its design, such as the grille and the headlights. The car's shiny exterior reflects the surroundings, adding depth to the image.  Throughout the video, the car is seen in motion, driving on a road, which adds a sense of speed and excitement to the video. The car's design and the way it moves are the main elements of the video, creating a dynamic and stylish showcase of the car's features."
+    ],
+    "video_ids": [
+        "aXug1j2kQPg_16_0to157"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by prominently featuring a red sports car. The car is clearly visible and occupies the central focus of the frames, matching the description. Additional elements like the background building and greenery do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features fulfills the 'Object(s)' condition as depicting featuring a red sports car in The car is shown visible in occupies a majority focus of each frames, with the description of The elements such the road and and theery do not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9e5c3e7aa7860ac286571d181b6305a684694a4d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57896a28ea14ade06813b0f935b97dccda48d77ddf520be06b6fa3d4fb30f510
+size 238141
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..6e28d5239d7bbc8e136f7f1c1714be98a8ad4786
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a basketball game between two teams, one wearing black and the other in white. The player in black is in the process of shooting the ball, his arm extended and the ball in mid-air. The player in white, wearing a headband and glasses, is in the act of defending, his arms raised and his body leaning forward in anticipation. The background is filled with the blurred figures of other players and spectators, adding to the intensity of the scene. The focus is on the two players, their actions and expressions conveying the high stakes of the game. The style of the video is realistic, capturing the action and tension of the moment with clarity and precision."
+    ],
+    "video_ids": [
+        "h960jd6knC0_35_0to155"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two players - one in black shooting the ball, another in white defending.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two basketball players as described: one in a black Georgetown jersey (number 15) preparing to shoot, and another in a white 'Orange' jersey (number 13) defending with his hand raised. The core description is accurately fulfilled, with no contradictory elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts two basketball players in described: one in a black jersey Hoy withnumber 33) attempting to shoot the and another in a white GeorgetownN' jersey (number 12) defending. his hands raised. The scene elements of fulfilled represented, with the additional elements present}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..33f98d4867733554070ca27b135e3b86583b1d79
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-115-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6aff0a4da0745c7cd3ffd8a6f88e19cc7dbdc55ea3538178961ca1bf03f0f020
+size 188612
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..6684e98e88a5a5b7e971280fe3e8e5f128a02da7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of koi fish swimming gracefully in a pond. The water is clear enough to see the bottom, which is covered with rocks and some aquatic plants. The koi fish display a variety of colors, including orange, white, and black, with some having intricate patterns on their bodies. They move fluidly through the water, occasionally creating ripples that disturb the surface. The camera remains stationary throughout the video, focusing on the gentle motion of the fish as they explore their environment. The overall atmosphere is peaceful, highlighting the natural beauty and tranquility of the pond setting."
+    ],
+    "video_ids": [
+        "c5357ca1986a633aecfb5bf7ab9de15ea2b87b165571677380500986f54705c0"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Koi fish? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows multiple koi fish swimming in a pond. Their distinctive orange, white, and black patterns are visible, and they are the main focus of the footage, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a koi fish swimming in a pond. The distinct colors, white, and black color are visible, and they are surrounded primary focus of the scene. fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..38025f6a92f855d7555a64deee2d323213ccfcdd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6e09a59153e8c777f6ca7d1f14a87b3e67d8c5003b2f148555d3c9245b1ada25
+size 235847
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..ba718a5d40c5c861b76677eb55376bdc13e2bf1e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene sunset scene at a beach. The sky is painted with vibrant hues of orange, yellow, and pink, transitioning into a darker blue as it meets the horizon. The sun is partially obscured by clouds, casting a warm, golden glow across the water. The reflection of the sun on the calm sea creates a shimmering path of light that stretches towards the shore. In the foreground, tree branches with green leaves hang down, framing the view and adding a natural border to the scene. A small dog can be seen walking along the sandy beach near the water's edge, moving from left to right across the frame. The overall atmosphere is peaceful and tranquil, with the gentle movement of the waves and the soft rustling of the leaves providing a soothing soundtrack to the visual beauty. There are no significant changes or movements in the scene throughout the video; it remains a static yet captivating depiction of a beautiful sunset at the beach."
+    ],
+    "video_ids": [
+        "6544b8bd13ab488431d72d974b80e1368a7e9edba3b76e2ddb491d1d7d5814ef"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Tree branches with green leaves, a small dog? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows tree branches with green leaves framing the top of the scene, and a small dog is visible on the left side of the beach near the water. These elements match the specified 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features shows tree branches with green leaves in the top left the scene, and a small dog appears visible in the right side of the frame, the water. These elements match the description 'Object(s)' condition without any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ec72124df851309ef33fe6e994f9c2f68ca96bf5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:632f280d038f99ba78954e53929d1ae4e6d48bea11cc14fcc3e7bc4fb9e3d05c
+size 195130
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..46648c3b75d4a3b0fab0d9a6b0ef8732d5ce2240
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man with a beard and a woman are engaged in a conversation in a barber shop. The man is gesturing with his hands, while the woman is leaning on the counter, listening attentively. The barber shop is well-stocked with various grooming products and tools, and the barber is standing behind the counter, observing the interaction. The scene is set in a cozy and inviting atmosphere, with warm lighting and comfortable furnishings. The video captures the essence of a typical day in a barber shop, with customers engaging in friendly conversations and the barber attentively observing the proceedings."
+    ],
+    "video_ids": [
+        "6gOdE_bxJT8_8_0to143"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard, a woman, and a barber standing behind the counter.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard, a woman, and a barber (a bald man in a white coat) standing behind a counter in what appears to be a barbershop or grooming store. The setting, characters, and their positions match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard standing a woman, and a barber standinglikely man man) a dark shirt) standing behind a counter. what appears to be a salonbershop or salon salon. The man includes the, and their positions align the description provided.}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..10dfdf9022a7f09631733bb75bdc46c57150f963
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d8ffa36529695958ad3046be04358c32e8f1397d6b95341d26eda088976836f4
+size 211166
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..e33de93078a090281ce41c099bc71993befaa1c2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen in a large, industrial warehouse. He is dressed in a bright orange jacket and a fur-lined hat, giving him a distinctive and eye-catching appearance. His hands are outstretched, as if he is gesturing or explaining something. The warehouse itself is spacious and filled with various items, including a red car parked in the background. The lighting in the warehouse is dim, with the man being the main focus of the scene. The overall style of the video suggests a narrative or story, with the man's actions and the setting hinting at a possible plot or theme."
+    ],
+    "video_ids": [
+        "Av8i1h9a6R0_45_24to151"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a bright orange jacket and fur-lined hat, a red car parked in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a bright orange jacket and a fur-lined hat, standing in what appears to be a parking garage. In the background, a red car is parked, matching the description. The scene is consistent with the specified 'Object(s)' condition without any conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing a bright orange jacket and a fur-lined hat. standing in a appears to be a large garage. In the background, there red car is parked, which the description provided The setting is consistent with the provided objectsObject(s)' condition.\"\n any contradictions elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ba0795e681a160a635266f477169781780420163
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:25e71ba597548411209959bcb0df2c9094b17c986c29570803ef854779a2fe53
+size 181101
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..acd61bbd9e365c993605bee66f1d1f27f90da77d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a sleek, black Volkswagen racing car on display in a showroom. The car is positioned in the center of the frame, with its front facing the camera. The showroom has a modern design, featuring a white floor and walls adorned with blue and white geometric patterns. On the wall behind the car, there's a large screen displaying a video of the car in action on a race track. The car is equipped with a large rear wing and a front splitter, indicating its aerodynamic design for high-speed racing. The overall style of the video is sleek and professional, highlighting the car's design and performance capabilities."
+    ],
+    "video_ids": [
+        "-VKul4hVjKs_13_0to154"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A sleek, black Volkswagen racing car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully features a sleek, black Volkswagen racing car, specifically the Volkswagen I.D. R Pikes Peak, which is clearly visible in both the outdoor track scene and the indoor exhibition setting. The car's design, branding, and color match the description, with no conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a sleek, black Volkswagen racing car. which a Volkswagen GT.D. R,ikes Peak, which is a visible in the frames foreground and scene and the indoor display setting. The car's design, including, and racing match the description provided and no conflicting elements present}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..13cc2ae1c30a8d3c760962e9d6cc445f183fedfa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4644b033d6466b6dbd0e80aac9a149e898c8e89895c008c9415f126c68966790
+size 179402
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..ee68f93ae701a2e7db618e06851084f730ba4713
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man holding a trophy, which is a large, clear glass award with a soccer ball on top. The man is wearing a blue and white outfit, and he is smiling as he holds the trophy. The trophy is significant in size, and it is clear that the man is proud of his achievement. The background is blurred, but it appears to be an indoor setting with a stage and a podium. The man is standing on the stage, and he is the main focus of the video. The trophy is the second most prominent object in the video, and it is clear that it is an important part of the event. The man's outfit is the third most prominent object in the video, and it is clear that he is dressed for a formal occasion. The man's smile is the fourth most prominent object in the video, and it is clear that he is happy and proud of his achievement. The stage and podium are the fifth most prominent objects in the video, and they provide context for the event. The background is the sixth most prominent object in the video, and it provides a sense of location and setting."
+    ],
+    "video_ids": [
+        "7XI0YcddG9o_69_58to213"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man, Trophy, Blue and white outfit, Smile? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man holding a trophy, wearing a blue and white outfit, and smiling. These elements match the 'Object(s)' condition described. The presence of another man and text overlay does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man holding a trophy, which a blue and white outfit, and smiling. The elements match the 'Object(s)' condition provided. The presence of the soccer in the in in not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7c20c13b371cd4e0c06178f43dcb977712702928
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d76ffdee6ab5e01a71321b3403b658896881c92ebdaebad15ef1f2db2ac3735b
+size 173798
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..f89960b1fe2ff3d6acbbcafda1c24cec8375d95a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a vibrant green Porsche sports car parked in front of a building with a large window. The car is positioned at an angle, showcasing its sleek design and black rims. The building's window reflects the car, adding depth to the scene. The car's shiny exterior reflects the surrounding environment, indicating that the video was taken during the day. The overall style of the video is dynamic and colorful, emphasizing the car's design and the urban setting."
+    ],
+    "video_ids": [
+        "mOXXjUpaFd8_61_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A vibrant green Porsche sports car, a building with a large window? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a vibrant green Porsche sports car, specifically a Porsche 911 GT3, parked outside a building with a large window. The car is clearly visible, and the building's facade with its large glass window is also evident in the background, matching the description. Additional elements, such as a person in a red hoodie and a poster of a race car, do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a vibrant green Porsche sports car, which a model 911 GT3, which in a building with large large window. The car's the the, and the building's large, a large window window is also evident in the background. fulfilling the description provided The elements like such as the person walking the yellow shirt walking a yellow on a car car, do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..75d54ab7d5b3db40baede60dfc4c6ae2c649317b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50fb1620d2f22bff261a2f7875f87e1eabd6a52fcb54fedbcd964cb3e474bb59
+size 192138
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..a11568a7cc81b2abacdbf517bf298b585576e3fa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a thrilling moment in a football game. The main focus is on a quarterback, who is in the process of throwing the ball. He is wearing a white jersey with the number 12 and the name \"Brady\" on the back. His arm is extended, and the football is clearly visible in his hand.  In the background, there are other players, some of whom are wearing green jerseys, indicating they are from the opposing team. The quarterback is on the field, which is a typical football field with its distinctive lines and markings.  The style of the video is dynamic and action-packed, capturing the intensity of the game. The camera angle is from the side, providing a clear view of the quarterback's throw and the surrounding action. The focus is on the quarterback and his throw, with the other players and the field serving as the backdrop to this exciting moment."
+    ],
+    "video_ids": [
+        "soLQhOrptkg_33_0to116"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A quarterback in a white jersey with number 12 and name 'Brady', other players from the opposing team, and a football.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a quarterback wearing a white jersey with the number 12 and the name 'Brady' on the back, consistent with Tom Brady of the New England Patriots. He is holding a football and appears to be in the process of throwing it. Other players from the opposing team (in green and white uniforms) are visible around him, and the scene is set on a football field, matching the described conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a football in a white jersey with the number 13 and the name 'Wady', on the back. which with the Brady's the New England Patriots. There is holding a football, appears to be in the process of throwing it. Other players from the opposing team,in green jerseys white uniforms) are visible in him, and the football is set on a football field. fulfilling the description conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f7157d247081fdcc566cddfa994a1c77b135d059
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1150-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee35c50f9f246bcd7b780152719a5847858c2b376eb5e6b168e5991a8892dc7f
+size 176520
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..39dfee1cc8a68a47359bb01b9a86d38a889a2c37
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen cooking in a kitchen. The person is using an orange pot on a gas stove. They are stirring the contents of the pot with a spoon. The kitchen is well-equipped with various utensils and ingredients. The person is focused on their task, ensuring that the food is cooked evenly. The overall atmosphere of the video is one of a homey and comfortable kitchen environment."
+    ],
+    "video_ids": [
+        "IqS9HuXm9iQ_42_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, an orange pot, a gas stove, a spoon? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person covering an orange pot on a gas stove. Although a spoon is not clearly visible in the frames, the presence of a wooden spoon on a tray nearby suggests it is part of the kitchen scene. The core elements (person, orange pot, gas stove) are clearly present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person stirring a orange pot on a gas stove with The the spoon is not visible visible, the video provided the person of a spoon spoon in the nearby in suggests that could part of the cooking setup. The video elements ofa, orange pot, gas stove) are present depicted and match the description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b6050c6bb7d12a0c8c0762fc89db108a06f4f547
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a27dfa79d15cccbaec12cb2e312b09143bb3144fa3d08dfc1f0ec7ce6397af6e
+size 216992
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..ad1fc52ef29dbc04d86186cb63ad620279065646
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two men are engaged in a conversation in a living room setting. The man on the left, wearing a plaid shirt, is gesturing with his hands as he speaks. The man on the right, dressed in a blue jacket, is listening intently with a surprised expression on his face. The room is comfortably furnished with a couch and a lamp, creating a relaxed atmosphere. The men appear to be deeply engrossed in their discussion, suggesting a serious or unexpected topic of conversation."
+    ],
+    "video_ids": [
+        "0imfrioKqPg_25_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men engaged in a conversation. One man is facing the camera, while the other is seen from behind. Their presence and interaction fulfill the 'Two men' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two men sitting in a conversation. They man is gest the other and gest the other is turned from the. The body and interaction fulfill the 'Object men' condition described any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3255596fbec708bbbd323e129f8c85f03627e124
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab249cbfad6a58960aeab3a19e7f509d498a137413bc5c8c60cc01ea6cabea87
+size 134234
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..99b2f208bb430929180aa45ad732f792340393bc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a football player in action, showcasing his athletic prowess and determination. The player, dressed in a white and blue uniform with the number 99, is seen in three distinct moments. In the first frame, he is seen in a defensive stance, his arms outstretched in a display of readiness. His eyes are focused, and his body language speaks of anticipation.  In the second frame, the player is seen in a dynamic pose, his body leaning forward as he prepares to make a move. His muscles are tensed, and his gaze is directed towards the field, indicating his readiness to spring into action.  The third frame captures the player in a moment of triumph. He is seen raising his arm in a victorious gesture, his face lit up with a smile. The background reveals a stadium filled with spectators, their faces a blur of excitement and anticipation. The player's triumphant pose suggests a successful play, adding to the overall narrative of the video.  Throughout the video, the player's uniform and the stadium in the background remain consistent, providing a sense of continuity. The video is a dynamic portrayal of a football player's journey, from anticipation to action to triumph."
+    ],
+    "video_ids": [
+        "-iIIpqbeSds_19_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A football player in a white and blue uniform with the number 99.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a white and blue uniform with the number 99 prominently displayed on the sleeve and chest. The player's attire, including the NFL logo and team name (Texans), matches the description. The background and lighting are consistent with a stadium setting, and there are no elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a white and blue uniform with the number 99 prominently displayed on the jersey. the. The player is attire and including the helmet-style and the colors,Cowans), align the description. The player shows setting are consistent with a football setting, which the are no additional that contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fb9c56ec6e8cdf972a379d67dd0d4e3f71bfbd6d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:df9c4b81ee3832a18b565f26eaae0f74f17e230e68e701dcac9f14a41c83b304
+size 274207
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..6ad188cedcdaf55af7a4928f7f93bb62a3c598dd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young woman with glasses, wearing a plaid shirt and a necklace. She is standing in a room with a brick wall and a refrigerator in the background. The woman is looking directly at the camera, and her expression is neutral. The lighting in the room is soft, and the overall atmosphere is calm and relaxed. The style of the video is casual and candid, capturing a simple, everyday moment."
+    ],
+    "video_ids": [
+        "J7JheWj56hk_10_44to256"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman with glasses, wearing a plaid shirt and a necklace. She is standing and looking directly at the camera with a neutral expression.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman wearing glasses, a plaid shirt, and a necklace, standing and looking directly at the camera. Her expression appears neutral, and the background elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman with glasses, a plaid shirt, and a necklace. standing and looking directly at the camera with Her expression is neutral, which the setting is, not contradict the description description.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ee19842ee069517a4d567c141f6515c47fddf62f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b9d00759d05a17617dbe8bca1bbc51508dd5f635205fd2f5655d4749094b00c4
+size 146890
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..90d3401ed6e80199d4883f64e3b33a9d076c711f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the transformation of a building facade from a state of disrepair to a vibrant, colorful display. The first frame shows a dilapidated brick building with peeling paint and a rusted metal structure. The second frame reveals the building being painted in a bright yellow color, with the addition of a Batman logo on the side. The third frame shows the completed transformation, with the building now adorned in a lively mix of colors and patterns, including a large, eye-catching mural. The style of the video is a time-lapse, showcasing the progression of the building's renovation over time."
+    ],
+    "video_ids": [
+        "UDHvfsMFPOY_18_0to119"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A dilapidated brick building, bright yellow paint, a Batman logo, and a large, eye-catching mural.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a brick building with visible signs of wear and age, consistent with 'dilapidated'. The walls are painted bright yellow, and a prominent Batman logo is displayed on the facade. Although the video does not show a large, eye-catching mural beyond the Batman logo, the core elements described are present, and additional elements like street signs and trees do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a building building that bright wear of wear and tear, which with adilapidated'. The building are painted in yellow, which there Batman Batman logo is displayed. the building. Additionally the video does not show a large, eye-catching mural, the yellow logo, the presence elements of in present. making the artistic like the lamps and a do not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..338025b54af0065b1f52640f44278b855e03d8ac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a2456e59fc09cadaf0c7b1bc14b59f8aa736733467c57358ef9355579351d83
+size 134141
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..7efaba108a110fed52ca96b3de136686aa628a4d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video begins with a view of a staircase leading upwards. The stairs have wooden treads and white risers, bordered by a white handrail on the right side. To the left of the stairs is a white door with horizontal slats and a brass doorknob. A bright yellow tennis ball rolls down the stairs, bouncing off the bottom step and landing on the wooden floor at the base of the stairs. The camera then pans to the right, revealing a light blue couch against a wall with a window covered by blinds. A dog, with a white and brown coat, is lying on the floor next to the couch, looking up at the camera. The dog appears relaxed and is positioned near the tennis ball. The camera focuses on the tennis ball resting on the wooden floor, emphasizing its bright color and smooth surface. The scene captures a playful moment involving the dog and the tennis ball in a cozy indoor setting."
+    ],
+    "video_ids": [
+        "947f36064391d8a74a9470717b85fc1bae5558f15ebd7a6e2be2555816187f6b"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ['Stairs', 'White door', 'Tennis ball', 'Couch', 'Dog']? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully includes all the specified objects: stairs are visible at the beginning, a white door is on the left, a tennis ball is shown bouncing and rolling, a couch is present with a dog resting on it, and the dog is clearly visible. All elements are present and match the description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful all the objects objects: St, visible in the top, a white door is seen the left side a tennis ball is on rolling on resting, a couch is present on a dog sitting on it. and a dog is also visible. The elements are consistent and do the description.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b3f8a0a38c0b274004ac994ee4ad30f89ba28881
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5d639e2aa946462193872f4e20f78ea8b23259a4e4bfc0a2df8a950fe26ed59
+size 213645
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..bfab10b974f3d2ea62291d6ba7da9ea97a1c94cc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a red car's headlight, focusing on the details of the light and the car's body. The car is parked in a garage, and the camera angle is from the front, slightly angled to the side. The headlight is bright and clear, with a white light that illuminates the surroundings. The car's body is shiny and red, reflecting the light from the headlight. The garage has a concrete floor and white walls, providing a neutral background that contrasts with the red car. The video is a still image, and there is no movement or action. The style of the video is a straightforward, clear, and detailed view of the car's headlight and body, with no additional elements or distractions."
+    ],
+    "video_ids": [
+        "B-9inXmMdX8_11_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red car with a headlight and car body.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red car, focusing on its headlight and the surrounding car body. The car's glossy red paint and headlight are clearly visible, matching the core description. Additional elements like the dealership background do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a red car with which on its headlight and part front car body. The car is shiny red paint and thelight design clearly visible, fulfilling the description description of There elements such the car setting do not contradict the core but}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e997b7d6d1290c9dd89e33e68708f11357c705f3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b2515fd9324eabba9f0d4c74b5a982fa350d8cf871796dd381280f7b9f31a5a
+size 104991
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..850254956f06da2698968387f28fe154dbed5ac7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a futuristic Audi car on display at a car show. The car is sleek and black, with a distinctive design that includes large, rugged tires and a unique rear window. The car is positioned in a well-lit showroom, with other cars visible in the background. The lighting highlights the car's design and features, making it the focal point of the scene. The overall style of the video is sleek and modern, with a focus on the car's design and features. The video captures the car's design and features in a way that highlights its futuristic and innovative design."
+    ],
+    "video_ids": [
+        "eSunj6WzO8I_58_30to195"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A sleek and black Audi car with large, rugged tires and a unique rear window.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a sleek, black Audi concept car with large, rugged tires and a unique rear window design featuring red LED lighting. The car is displayed in an auto show setting, which does not contradict the core description. Additional elements like the background branding and people are acceptable as they do not conflict with the primary object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a sleek and black car car car with large, rugged tires and a unique rear window.. a ta lights. The car is displayed in a indoor show setting, which align not contradict the description description. The elements like the showroom and and the are present as they do not conflict with the main focus description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8a7e898d181587cd06f5f5f52e59594ef9707efc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1151-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ef4f1e6034aecec3c87e7dd4cb28b982823f9c03a6c85115d15f5d50a697053
+size 142678
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..0ae06b299745180ea453dae2c83208f887a3ba86
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a tender moment between two chimpanzees in a zoo enclosure. The chimps, with their dark fur and expressive faces, are seen in close proximity to each other, their faces pressed together in a kiss. The metal fence of their enclosure forms a grid-like pattern in the foreground, adding a sense of depth and perspective to the scene. The background is filled with lush greenery, providing a naturalistic backdrop to the chimps' interaction. The video is shot in a realistic style, capturing the chimps' behavior and the environment of their enclosure with clarity and detail."
+    ],
+    "video_ids": [
+        "KNZmmdwZ5iY_9_22to145"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two chimpanzees? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two chimpanzees positioned behind a wire mesh fence, interacting with each other. Their physical features, posture, and proximity to each other align with the description of two chimpanzees. There are no elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two chimpanzees interacting close a fence fence,. engaging with each other. The physical features and such, and the to each other align with the description of two chimpanzees. The are no additional in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3386278f7f85bd8f889d61b1ffb9d1daadcf6e05
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c0168fe58a7f295f14cccca22b09dd8a6506908aa466e016de9aeb91b130187
+size 261395
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..12145bfaf54d5eddffb369cace0b1f8195a963c8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic scene of a motorcycle rider performing a burnout on a wet street. The rider, dressed in casual attire, is seen in the center of the frame, skillfully maneuvering the motorcycle. The motorcycle, a sleek black model, is in motion, with smoke billowing from its rear tire, indicating the intense friction between the tire and the wet road. The rider's focus and control are evident as they maintain their balance on the motorcycle. The background features a grassy area, adding a touch of nature to the urban setting. The overall style of the video is action-packed and dynamic, capturing the thrill and excitement of the motorcycle burnout."
+    ],
+    "video_ids": [
+        "sUFNeOGAlTI_18_0to167"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Motorcycle rider, motorcycle? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a motorcycle with a rider standing next to it. The rider is partially visible, and the motorcycle is prominently featured, matching the 'Object(s)' condition. Additional elements like another person and smoke are present but do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a motorcycle rider a rider, on to it, The rider is wearing obscured, and the motorcycle is in featured in fulfilling the descriptionObject(s)' condition. The elements like the person in a are not but do not contradict the core description.\"\n}<|im_end|>\n addCriterion",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..65248adf004439da6279ce840eb76262a8f7ac2c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4152f983663711b4198e54570cab11fb64603241e574eb0fba93413f6171d086
+size 196928
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..8ba02ff672d477f9686b1e2dcb6390d549cac8f3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young girl with brown hair and a blue and white striped shirt is seen interacting with a large, friendly-looking man with a white beard and a purple plaid shirt. The man is wearing blue overalls and a straw hat. They are standing in a colorful, cartoon-like environment that includes a windmill, a red and yellow striped tent, and a green lizard. The girl is holding the man's hand, and they appear to be walking together. The overall style of the video is whimsical and playful, with vibrant colors and a cartoonish aesthetic."
+    ],
+    "video_ids": [
+        "OP5jKcylMxY_104_0to148"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young girl with brown hair and a blue and white striped shirt, a large, friendly-looking man with a white beard and a purple plaid shirt wearing blue overalls and a straw hat.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a young girl with brown hair tied in a ponytail, wearing a blue and white striped shirt, holding hands with a large, friendly-looking man. The man has a white beard, is wearing a purple plaid shirt, blue overalls, and a straw hat. All core elements of the description are accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a young girl with brown hair wearing in a ponytail, wearing a blue and white striped shirt and and hands with a large man friendly-looking man. The man has a white beard, a wearing a straw plaid shirt, blue overalls, and a straw hat. The these elements of the description are present represented in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7456e6f016eed3b789abd3fec39f39635bcfb250
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a4d6d374885bd478df67ff683adefadf4cec8a4cba83da460354ca149076a67b
+size 283805
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..f6dce8214cee0e860ce86574caaccdef5a0a5b7a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a basketball player in action during a game. The player, dressed in a white and blue striped jersey with the word \"Charlotte\" on it, is holding a basketball in his right hand. He is seen pointing with his left hand, possibly directing his teammates or strategizing his next move. The background is filled with a crowd of spectators, indicating that the game is taking place in a large arena. The player's focused expression and the dynamic nature of the scene suggest a high-stakes moment in the game."
+    ],
+    "video_ids": [
+        "0ShS3Q6XedQ_65_43to193"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A basketball player, a basketball, spectators? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a basketball player in a Charlotte jersey holding a basketball and pointing, with spectators visible in the blurred background. All core elements described in the condition are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a basketball player in a white Hornets holding a basketball, appears, which spectators in in the background background. The the elements of in the condition are present: any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d1bcac0b876f81d9bbfc08d1c5d91e5ba59a7f8a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c1fc0717d4c33c78a005581ef9e7b0fdc9392a37fee4475f1ea4fd9f6f934376
+size 204210
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..0d271cabebbe32e6466a93c3742af9363622a20d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with blonde hair, wearing a red shirt, and a braid. She is standing in a room with blue storage bins and shelves in the background. The woman appears to be speaking or reacting to something, as she has her mouth open and her eyes are focused on something off-camera. The style of the video is a close-up shot of the woman, with the background slightly blurred, emphasizing her as the main subject. The lighting in the room is bright, and the overall atmosphere seems to be casual and relaxed."
+    ],
+    "video_ids": [
+        "CX2Yli5okR8_289_0to155"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with blonde hair, wearing a red shirt, and a braid.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair styled in a braid, wearing a red shirt, which matches the core description. The background elements do not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair, in a braid, wearing a red shirt. which matches the description description provided The presence and, not contradict the description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1f985d09ae5df51a9de0519f40bbdda2f3fecc19
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5351cb8490a2702163f1655d8ab35f2df7d9f75d705aa8f0b9e5cc63ae084f2
+size 170280
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..0677227b56e1bc37e3876199ff28033f3b99f871
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a white Hyundai car driving down a snowy road. The car is sleek and modern, with a distinctive front grille and headlights. The road is covered in a layer of snow, and the car's tires leave tracks in the snow as it moves. The car is the main focus of the video, and it is captured from a side angle, allowing viewers to see its design and features. The background is a mix of snow-covered trees and a clear sky, creating a serene and peaceful atmosphere. The car's license plate is visible, and it reads \"AN-10497\". The video is likely a promotional or advertisement video for the Hyundai car, showcasing its performance and design in a winter setting."
+    ],
+    "video_ids": [
+        "rexE3Vx-A8k_0_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white Hyundai car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a white Hyundai car, which matches the specified object. The car is prominently featured throughout the video, and no elements contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a white Hyundai car, which matches the description '. The car is prominently displayed in the frames, and its other contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4baeb512992992308d6a45e11cc26389089fc59f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2c0190270296982e5ec7580e1017619cd5d3d80118882175a417e932e4ed4e0
+size 256192
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..e8e89e890210b1636b4d47e928346665de737012
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a kitchen, preparing food. He is wearing a black apron with the Guinness logo on it. The kitchen is well-equipped with a sink, an oven, and a window. The man is using a knife to cut a piece of food on a cutting board. The video captures the man's actions in a realistic and detailed manner, providing a clear view of the kitchen and the man's activities. The style of the video is realistic and informative, likely aimed at teaching viewers how to prepare a specific dish."
+    ],
+    "video_ids": [
+        "J-o-JTJTa4k_8_0to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a cutting board, a knife, a piece of food, a sink, an oven, and a window.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing an apron, standing at a kitchen counter with a cutting board, a knife, and a piece of food (an onion) on the board. In the background, a sink and an oven are visible, along with a window. All specified objects are present and accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing an apron, standing at a kitchen counter with a cutting board, knife knife, and a piece of food (which un). being the board. There the background, there sink and a oven are visible, and with a window. The the objects are present and the depicted in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..15e649773c1b4eeb9f74aa3c66ee3008eefae9ef
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7045894ff2d0661af19b276b7128d86379c13d2e38da76b7c787764d09115de6
+size 155739
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..444d0311a4e394844b587d1cbef13993ae98f0fe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video begins with a close-up view of a clear plastic bag containing a small, dark object. The camera slowly pans across the surface of the bag, revealing its texture and the way light reflects off its smooth, slightly wrinkled material. As the camera moves, it captures the subtle interplay of light and shadow on the bag's surface, highlighting the intricate details of the plastic's texture.  The camera then shifts focus to a different section of the bag, where a small, dark object is visible. This object appears to be a piece of debris or a small insect, partially obscured by the plastic. The camera zooms in on this object, providing a closer look at its shape and texture. The lighting continues to play a significant role, casting a rainbow-like reflection on the plastic around the object, adding a dynamic visual element to the scene.  As the camera continues to pan, it reveals more of the bag's contents, which include various small particles and possibly some liquid. The camera movement is"
+    ],
+    "video_ids": [
+        "a000f0e1de7ef5821482d8b08215ef16e45841dab75e18f9753c384c65a57e40"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Clear plastic bag, small dark object (debris/insect), small particles, possibly liquid inside the bag.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a clear plastic bag with visible wrinkles and reflections, containing small dark specks that could be debris or an insect, along with small particles and what appears to be liquid or condensation inside. These elements align with the described conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts a close, bag with a reflections and reflections, which a dark objectscks that resemble be interpreted or an insect. and with small particles that possibly appears to be liquid inside aensation inside the The elements align with the description ',}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b2b1d136351c34c80192e5cd0c660bda2ba77b81
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1152-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9138c658da7fc99eca6349f6c2da71a752bf1fa02e1886fda4f3e597a00c77e1
+size 268384
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..7e5f4e7e4473eee54a26e66e8591877efaeef22e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated depiction of a medieval king standing in a grand hall. The king is dressed in a red and gold armor, complete with a crown on his head. He is standing on a checkered floor, with a large archway in the background. The king's expression is serious, suggesting a moment of decision or contemplation. The overall style of the video is realistic, with attention to detail in the armor and the grand hall. The lighting and shadows add depth to the scene, creating a sense of realism. The video captures the majesty and power of the king, as well as the grandeur of the setting."
+    ],
+    "video_ids": [
+        "J-eN4QSMz0o_17_0to120"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A medieval king wearing red and gold armor with a crown.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a medieval king wearing red and gold armor with a crown, matching the core description. The armor is detailed with red fabric and gold accents, and the crown is clearly visible on his head. Additional elements like the background setting and other figures do not contradict the primary description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a character king wearing red and gold armor with a crown. which the description description. The attire is detailed with gold and and gold accents, and the king is prominent visible. the head. The elements like the check and do the details are not contradict the main description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..19a9969f352b3388e7abed1037c60b42c1d882ac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7ecd6a76e8593a2a3e8f43b6a44c230640ed2aef8f35a52242f5e89fd427baab
+size 193352
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..59cef6b1cb3b9f6e20f494a9d6e1667db72db030
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a kitchen, preparing a meal. She is wearing a red apron and a white shirt. The kitchen is well-equipped with various appliances and utensils. In the first frame, she is seen holding a bowl, which she places on the counter. In the second frame, she is seen pouring something from a bottle into the bowl. In the third frame, she is seen holding a knife, presumably to cut something. The kitchen has a microwave, an oven, and a sink. There are also several bottles and a vase on the counter. The woman appears to be in the middle of cooking, and the video captures her actions in a realistic and detailed manner."
+    ],
+    "video_ids": [
+        "0Qp3Ms6IWos_5_44to205"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a bowl, a bottle, a knife? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a kitchen setting, wearing an apron and standing near a counter with various items. A black cast iron skillet (which can be considered a type of bowl) is visible on the counter, along with a green bottle (likely dish soap or hand soap) and a knife block containing multiple knives. All specified objects are present in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a kitchen setting. interacting an apron, interacting at a counter. various kitchen. She bowl bowl- skillet iswhich can be considered a bowl of bowl) is visible on the counter. and with a bottle bottle andwhich a soap) a sanitizer) and a knife.. a knives. The the objects are present in the scene,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dfdb9a2bb77f8ab9bf97002c709ddc71f098ea6c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:66d11ef8e779140677a74d7ccd06a2591823d0d5b25430142277eec610cfe39c
+size 134311
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..4dd149f7fa29ccc01440c63042e75d83b5801c5a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a formal meeting taking place in a conference room. The room is adorned with a large wooden table, around which five individuals are seated. The participants are dressed in business attire, with some of them wearing ties. The table is cluttered with various documents and papers, indicating an ongoing discussion or negotiation. The room is decorated with a large flag, adding a touch of national pride to the setting. The overall atmosphere suggests a serious and professional environment."
+    ],
+    "video_ids": [
+        "CX5YvKMqWNM_48_24to205"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Five individuals seated around a large wooden table. Various documents and papers are on the table.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows five individuals seated around a large wooden table, with various documents and papers spread across the table. The scene is consistent with the described condition, and no elements contradict it.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows five individuals seated around a large wooden table. with various documents and papers placed out the table. The setting is set with the description ', and there additional contradict the.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3fd31d04f84ac19084509dddfec53c3e5d3ad7c4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e22a6a6e650e39c16c817fc3d90451d7a2ed380bf70ec28c6532805a4c703252
+size 154513
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..18ad45c35f8dd710e65b505cc6f65cb2d507bf43
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up shot of a tree with a variety of fruits growing on it. The fruits are in different stages of ripeness, with some being green and others yellow. The tree is lush and full, with a dense canopy of leaves. The fruits are hanging from the branches, and the tree appears to be healthy and thriving. The video is likely educational, showcasing the growth and ripening process of these fruits. The style of the video is straightforward and informative, focusing on the natural beauty and growth of the tree and its fruits."
+    ],
+    "video_ids": [
+        "mAtw7rN7fco_48_22to172"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A variety of fruits growing on the tree, in different stages of ripeness (green and yellow).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows papayas on a tree in various stages of ripeness: some are green (unripe), while one prominently displayed fruit is yellow (ripe). This matches the condition of a variety of fruits growing on the tree, in different stages of ripeness.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts aayas growing a tree, various stages of ripeness, some are green,unripe), and others is displayed is is yellow (ripe). This matches the description of having variety of fruits growing on the tree, in different stages of ripeness (}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..852cc16d468bf06cd0736525ca9f209c0e68e736
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e673b70ca62a0ca8aef1672c349814efb942205d34402624216f2c7cc2ef4830
+size 163954
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..09b5549c02804648065e054228afe26823a75e5e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young woman with long brown hair, standing in a kitchen. She is wearing a black top and is smiling at the camera. The kitchen is well-equipped with various appliances such as a microwave, oven, and coffee maker. There are also some items on the counter, including a bottle and a cup. The woman appears to be in a good mood, possibly preparing to cook or bake something. The overall style of the video is casual and friendly, with a focus on the woman and her surroundings."
+    ],
+    "video_ids": [
+        "75Zhx4zRCG4_0_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman with long brown hair, wearing a black top, and smiling at the camera.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman with long brown hair wearing a black top, and she is smiling at the camera. The background elements (kitchen appliances, cabinets, etc.) do not contradict the core description and are acceptable as additional context.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young woman with long brown hair, a black top, and she is smiling at the camera. The setting includes,kitchen setting, utens, and.) do not contradict the description description and are acceptable as additional elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..927a44832267cafe2d072f991b1e9d503a8af11f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d1a0fcb37bf69f8e436cece2a1228a87736f4147bf2c8e545506cb6e55713aa
+size 151220
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..4e5cf9fe37fa3b46034b6fbf4801b973773aedaf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the process of cooking a dish in a large pot. The pot is filled with a bubbling sauce, and various ingredients are being added to it. The ingredients include meat, vegetables, and herbs. The pot is placed on a stove, and the heat is turned on, causing the sauce to boil. The steam rising from the pot indicates that the dish is being cooked at a high temperature. The style of the video is a close-up shot of the cooking process, focusing on the pot and the ingredients being added. The video does not contain any text or narration. The overall impression is that of a home cooking scene, with the cook carefully preparing a delicious meal."
+    ],
+    "video_ids": [
+        "zHvKw8khHm0_4_96to251"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Large pot, bubbling sauce, meat, vegetables, herbs? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large pot with bubbling sauce, containing visible pieces of meat and what appear to be vegetables and herbs. The description is largely accurate, with no significant contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large pot with abling sauce, which what pieces of meat, vegetables appear to be vegetables and herbs. The presence of largely matched, with the significant contradictions or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6ec8b60068ecec6240399b9273a76edb7ef3d7b2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd1507b11cacf7ad62e733eb367501442ef73422035752760355da784ee9d954
+size 208783
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..d53a22325d1129d38b7327efe3b95ee9cd37c5d3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with short black hair, wearing a blue and white patterned blouse. She is standing in a room with a white wall behind her. The woman appears to be speaking, as suggested by her open mouth and focused expression. The lighting in the room is bright, and the overall style of the video is simple and straightforward, with no additional objects or background details to note."
+    ],
+    "video_ids": [
+        "W5Pe1zdo4o4_9_0to103"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with short black hair wearing a blue and white patterned blouse, standing and speaking.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with short black hair wearing a blue and white patterned blouse, standing and speaking. Her facial expressions and mouth movements indicate she is talking, and the overall scene matches the description without contradicting it.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with short black hair wearing a blue and white patterned blouse. standing and speaking. The attire expressions and the movements suggest that is speaking, which the setting appearance align the description provided anying it.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5f00a8988c4bab3da37be37ff59652ee84a460fe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0fcc7358c8f602cc578c86bfecdd7804aa14eef2793d37c718480270f65fc3be
+size 352253
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..f254cd53a5b13a2a90ffa10af01f0f9d0f53562d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a wooden dresser with three drawers, placed outdoors on a gravel surface. On top of the dresser, there are three figurines: a teddy bear, a Darth Vader action figure, and a Stormtrooper action figure. The teddy bear is positioned on the left side of the dresser, while the Darth Vader and Stormtrooper figures are on the right side. The dresser is situated in front of a white house with a gray roof, and there is a wooden fence to the right of the dresser. The style of the video is a simple, straightforward depiction of the scene, with no additional action or movement. The focus is on the dresser and the figurines, with the background providing context for the setting."
+    ],
+    "video_ids": [
+        "cWM2GsTa3LA_79_55to218"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A wooden dresser, three figurines (a teddy bear, a Darth Vader action figure, and a Stormtrooper action figure)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a wooden dresser with three figurines on top: a teddy bear (which resembles Chewbacca), a Darth Vader action figure, and a Stormtrooper action figure. These elements match the description exactly, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a wooden dresser with three figurines on top: a teddy bear,which resembles abacca from a Darth Vader action figure, and a Stormtrooper action figure. The elements match the description provided, and there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..82692e2b5b1cc19fc0b581aaf309e5d36cc249b2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1153-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:93367c84d658cb151512c609a154565cad1013b112dec4b765f8d84d7ff5526b
+size 227492
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..79319e20d4736e745c8be3767e97502a93d27e7a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a blue Hyundai SUV parked in a parking lot. The car is positioned at an angle to the camera, allowing a clear view of its front and side. The Hyundai logo is prominently displayed on the grille. The car is parked next to a fence, and there are mountains visible in the background. The lighting suggests it is daytime. The style of the video is a straightforward, real-world depiction of the car, with no additional effects or embellishments."
+    ],
+    "video_ids": [
+        "0c_QYnepf6k_10_0to179"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue Hyundai SUV parked at an angle, with its front and side clearly visible, and the Hyundai logo on the grille.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a blue Hyundai SUV parked at an angle, with both its front and side clearly visible. The Hyundai logo is prominently displayed on the grille, matching the description. Additional elements like the license plate and background scenery do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a blue Hyundai SUV parked at an angle, with its its front and side clearly visible. The Hyundai logo is prominently displayed on the grille, matching the description. The elements like the mountain plate and the scenery do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fe9720c99dfe493448162cc5966b84413b9bcff1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6df6bad56091aa053e74f7375de45a62a807614060b62dc5cbacf3cfde8fb071
+size 250613
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..a0eb23435e791cf74c7a082eeae12ade11642055
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man named Peter McLaughlin, a resident, who is being interviewed on a news channel. He is wearing glasses and a blue shirt. The background of the video shows a brick wall with a red cross sign. The news channel is tracking Dorian, a storm, and the man is likely discussing his experiences or concerns related to the storm. The style of the video is a standard news interview, with the interviewee in the foreground and the news ticker and logo in the background."
+    ],
+    "video_ids": [
+        "aI41ibgz2LQ_13_0to157"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man (Peter McLaughlin), glasses, blue shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man identified as Peter McLaughlin, wearing glasses and a blue shirt, which matches the specified 'Object(s)' condition. Additional elements such as the brick wall, Red Cross sign, and news graphics do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing as Peter McLaughlin, wearing glasses and a blue shirt. which align the description 'Object(s)' condition. The elements such as the background wall and green Cross symbol, and the ticker are not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..37812d56eed6e9bb3b72612e8c8d10ac1d2421a5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e2c9178e975de581f2cedc0cd68e6604365d82f8f266c74f43955fcf49d83ca
+size 154779
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..642ea3ac486dd420c06ff7a45e689e00c9b145e3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man in a plaid shirt and overalls, wearing a cowboy hat, interacts with two children in a rural setting. The man is holding a red cup and appears to be pointing at something. The children, a boy and a girl, are standing next to a blue bucket. They are looking at the man with interest. The background features a wooden bench and a chair, suggesting a relaxed, outdoor setting. The overall style of the video is cartoonish and colorful, with a focus on the characters and their interaction."
+    ],
+    "video_ids": [
+        "2oopbQOW7WM_334_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a plaid shirt and overalls with a cowboy hat, a boy, a girl, a red cup, and a blue bucket.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a plaid shirt and overalls with a cowboy hat, a boy, and a girl. A blue bucket is visible on the ground, and the man is holding a red can labeled 'BIG FAT WORMS', which can be considered a red cup in context. All specified objects are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man wearing a plaid shirt, overalls with a cowboy hat, standing boy, and a girl. The red bucket and visible, the ground, and the man is holding a red cup. 'C'. SORMS'. which could be interpreted a red cup. this. The elements elements are present in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7201faab5db6af95c44b0eb0c7fcbeec80633023
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:abb37deace807a3f4940e81ec4652b669dc5b06cbbed6a48a0ac0256394392d9
+size 162876
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..d5df90339b7cdbcabfa772feedb3cccb419077ce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in a field with trees and bushes in the background. He is wearing glasses and a leather jacket. The man appears to be speaking, suggesting that he might be giving an interview or a speech. The setting is outdoors, and the man is standing in front of a pile of rocks. The overall style of the video is natural and candid, capturing a moment in the man's life."
+    ],
+    "video_ids": [
+        "UZdl_-KI5Jw_13_0to189"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man wearing glasses and a leather jacket, standing and speaking.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing glasses and a leather jacket, standing and speaking. The background elements (trees, ground, etc.) do not contradict the core description and are consistent with an outdoor setting. The man's attire and action match the specified condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a man wearing glasses and a jacket jacket, standing and speaking. The man includes,green and grass) and.) do not contradict the description description. are acceptable with the outdoor setting. The man's attire and posture of the description conditions.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c86c293e1dbbb005fb677b91b1702d9da73f1f2a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5f95715d35822a300580721cc4329aa967e2ed3091a4d0b89f8cca91249591e
+size 168131
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..e3e24baeb99ef1b2598a54b86f62e3c590b36505
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a television interview featuring two men in a studio setting. The first man is seated on a black chair, wearing a dark suit and tie, and appears to be speaking or listening intently. The second man is seated on a gray couch, wearing a dark suit and tie as well, and seems to be engaged in a conversation with the first man. Between them is a small round table with a glass top, on which there are two cups. The background of the studio includes a large screen displaying the logo of the television channel, and there are multiple monitors and a bookshelf visible. The style of the video is professional and formal, typical of a news or current affairs program."
+    ],
+    "video_ids": [
+        "MMsHeje0N5E_9_24to239"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, a black chair, a gray couch, a small round table with a glass top, two cups, a large screen, and a bookshelf.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing two men, a black chair, a gray couch, a small round table with a glass top, two cups, a large screen, and a bookshelf. All these elements are present and accurately depicted in the scene without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing two men seated a black chair, a gray couch, a small round table with a glass top, two cups, a large screen, and a bookshelf. The these elements are present in match depicted in the scene.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fb1d6732c3b2a91a14ce263b81e1dca01c2a105f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc4a68c331b0b87f784017e3d3eca7a0576e46006fbc11f480af86661dfceb9f
+size 119646
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..e14aa9fb731265c9698c3d1fa4a0b7fe1774813a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up of a tiger's face in a natural setting. The tiger's eyes are closed, and its mouth is slightly open, revealing its teeth. The tiger's fur is a mix of orange and black stripes, and its whiskers are white. The background is blurred, but it appears to be a grassy area with trees. The style of the video is realistic, with a focus on the tiger's facial features and expressions. The camera angle is close-up, allowing the viewer to see the details of the tiger's face. The lighting in the video is natural, suggesting that it was taken during the day. The overall mood of the video is calm and serene, with the tiger appearing to be at ease in its natural habitat."
+    ],
+    "video_ids": [
+        "RQliXWIDk_w_104_0to139"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A tiger's face with closed eyes and slightly open mouth revealing teeth, orange and black striped fur, and white whiskers.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a tiger's face with its eyes mostly closed and mouth slightly open, revealing teeth. The tiger has orange and black striped fur and prominent white whiskers, matching the description. The background is blurred, focusing attention on the tiger's facial features, which are consistent with the requested object details.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a close-up of a tiger's face with its eyes closed closed, its slightly open, revealing teeth. The tiger's orange and black striped fur, white white whiskers. which the description provided The background is blurred, focusing attention on the tiger's face features, which is consistent with the given object.\"\n.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ae5c5e019bd38777e9469c706275246b6e3d0d94
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b26775baee2a168d2b99416322bdd85b899cd7ebbe2980609c16828b1981060f
+size 275233
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..bb15ca7afc23d85f60a43aa550e9913eec5d367a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene mountain landscape, where a river flows through a rocky valley. The river, with its clear blue waters, is surrounded by large boulders and lush green trees. The camera follows the river's path, providing a bird's eye view of the landscape. The sunlight filters through the trees, casting a warm glow on the scene. The overall style of the video is naturalistic, capturing the beauty of the wilderness in its raw form."
+    ],
+    "video_ids": [
+        "vZWFrG9uSec_28_0to145"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: River, large boulders, lush green trees? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a river flowing through large boulders, surrounded by lush green trees. The scene is consistent with the described elements without any contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a river flowing through a boulders and surrounded by lush green trees. The vibrant is vibrant with the description elements, any contradictions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3f5005905c0353ee2e0e163df1b7a6a5ac186d5f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:afe6598f09f2402764df51f211c56c57c9c566a20f00c1487c874e6096bfd095
+size 261366
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..9848992a70c9dea98bc55e2cd8afde22e829000b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man with short dark hair is seen in a kitchen setting. He is wearing a green shirt and appears to be in deep thought or possibly experiencing a headache. He is holding his hand up to his ear, pressing it against his head. The kitchen has wooden cabinets and a white countertop, and there is a window in the background. The man's expression is serious and contemplative. The overall style of the video is realistic and it seems to be a scene from a drama or a movie."
+    ],
+    "video_ids": [
+        "6lXojEPj-iM_5_44to214"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with short dark hair, wearing a green shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with short dark hair and a green shirt, which matches the description. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man with short dark hair wearing he green shirt, which matches the description provided The setting appears additional elements in not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..02041d69fe404d739f8982366295aca8cdea7ed7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1154-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cfaec595650d3fd1dc35078e240c3207385d013ff8cba147e81957fefa572085
+size 118123
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..0bb36cda375af3680afad25c4bf4bfa98ec67280
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a vintage car in a showroom setting. The car is black and white, with a shiny exterior that reflects the showroom's interior. The car's interior is visible, featuring a white leather seat and a silver steering wheel. The steering wheel is adorned with a logo, indicating the car's brand. The car is parked in a showroom, which is filled with other cars in the background. The showroom has a polished floor that reflects the cars and the lights. The video captures the elegance and sophistication of the vintage car, highlighting its design and craftsmanship."
+    ],
+    "video_ids": [
+        "WciLAb4Ni4s_16_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black and white vintage car, a white leather seat, a silver steering wheel with a logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a vintage car with a black exterior and white interior, including white leather seats. The steering wheel is silver and features a central logo, matching the description. The presence of other cars in the background does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a black car with a black exterior and a interior, which a leather seats and The steering wheel is silver with features a logo logo, which the description. The video of additional cars in the background does not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..099a49d180011882b2dff5d1b6abd03095726377
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f45d39ecdee5f5edc54d908e2e2d3a499dd901cd3d5eb0476b36c1b7065d7743
+size 148364
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..aa26d19d6f3e2c83cc6de3bb152c61032892a41f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a red and white toy mask with yellow eyes, resembling a character from a popular franchise. The mask is placed on a black pole, and it appears to be in motion, as if it's moving towards the camera. The background is filled with various comic book covers, suggesting a theme of pop culture and fandom. The style of the video is playful and lighthearted, with a focus on the toy mask as the main subject. The comic book covers in the background add a sense of nostalgia and excitement, hinting at the broader world of fan culture and collectibles."
+    ],
+    "video_ids": [
+        "yxsn0Y2T0-Y_19_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red and white toy mask with yellow eyes? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red and white toy mask with prominent yellow eyes, which matches the description. The mask is clearly visible and the background elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red and white toy mask with yellow yellow eyes, which matches the description provided The mask is the the and the colors,, not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4829a185fe75092a4dec8d83ca122594aa62e79b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a3ce043605898eb94860bf3de5c16f1d1c7cd8d36c7b531a529e9296ef17d3f3
+size 172636
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..d11e9d45b3c6697588b5a1a40ed7b3537b3fb5aa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen preparing a meal on a wooden table. The person is using a knife to slice sausages and place them on a white plate. The sausages are accompanied by a green sauce and red tomatoes, adding a pop of color to the dish. The table also holds a pink bowl and a blue plate, both of which are empty. The person is dressed in a blue shirt, which contrasts with the warm tones of the wooden table. The overall style of the video is casual and homey, capturing the simple joy of cooking."
+    ],
+    "video_ids": [
+        "UmemVqN8DSE_82_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person, knife, sausages, white plate, green sauce, red tomatoes, pink bowl, blue plate, blue shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a person garnishing a plate of sausages with red tomatoes. The white plate, green sauce, pink bowl, blue plate, and blue shirt are all visible. Although a knife is not present, the core objects described are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a person inishing a plate of sausages with green tomatoes and The person plate, green sauce, and bowl, and plate, and blue shirt are all present in The the knife is present being in the action elements ( in all represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..83bb0fe6bc0a19813568a2a188e4b0f7fe0f86dc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88813b718f8dd2e6c5ee0c987be4986c8c5342fd349d1e55c9b8b899ce43c10b
+size 131120
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..88268125c41b3bb4e0202f66bec5b8074b0d33b2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a triumphant moment in a mixed martial arts (MMA) match. A female fighter, wearing a black tank top with the UFC logo, stands in the center of the octagon, her arms raised in victory. Her face is lit up with a smile, reflecting her joy and accomplishment. The referee, dressed in a black shirt, stands next to her, his hands raised in a gesture of congratulations. The background is a blur of red and white, suggesting the presence of a crowd and the excitement of the event. The video is shot from a low angle, emphasizing the fighter's victory pose and the referee's congratulatory gesture. The overall style of the video is dynamic and action-packed, capturing the intensity and excitement of the sport."
+    ],
+    "video_ids": [
+        "8iK1n1P3JMY_17_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A female fighter and a referee.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a female fighter in a UFC outfit with her arms raised, and a referee in a black shirt and gloves raising her hand, which matches the 'Object(s)' condition. Additional elements like a cameraman are present but do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a female fighter in a UFC uniform, a arms raised, which a referee is a black suit is white is her hand, indicating is the descriptionObject(s)' condition of The elements like the crowdaman and present but do not conflict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..22d9701cfaf24f3e1de5cb3e97b8f8f7192a34cf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3bcdb40fd4e1ec14eb26c3ed6cfb3e0acd6f90eac6a97b332657f6a896c7c813
+size 296998
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..473b3f6bd620544dadca494d3493e16f1b7920ea
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with curly hair and large sunglasses, standing on a tennis court. She is wearing a gray shirt and appears to be in a thoughtful or contemplative mood. The tennis court is green and well-maintained, with a clear sky above. The woman's sunglasses are prominent, adding a touch of style to her appearance. The overall atmosphere of the video is calm and serene, with the focus being on the woman and her surroundings."
+    ],
+    "video_ids": [
+        "BFDkJzmhfEw_8_53to186"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with curly hair and large sunglasses, wearing a gray shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with curly hair and large sunglasses, wearing a gray shirt, which matches the description. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a woman with curly hair and large sunglasses, wearing a gray shirt. which matches the description provided The background appears additional elements are not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..982a9a1419dd34984451283c07d3d2f3cd47569e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:19f780e47034b5d0cedb674fd1fa5db451b20421a2905559bf16abda237d00be
+size 143138
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..689f4d7634a7964a43c9b0060528e4347f2c4268
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men engaged in a conversation. The first man is wearing a black jacket and has a serious expression on his face. The second man is wearing glasses and a black sweater, and he appears to be listening intently to the first man. They are standing on a porch with a white railing, and there is a tree in the background. The lighting suggests it is daytime. The style of the video is realistic and it seems to be a scene from a movie or TV show."
+    ],
+    "video_ids": [
+        "Z4zb5uepYw8_28_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men are present. The first man wears a black jacket and has a serious expression. The second man wears glasses and a black sweater, and he is listening intently.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men. The first man is wearing a dark jacket (appears black) and has a serious expression. The second man is wearing glasses and a black sweater, and he appears to be listening intently. These details align with the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men standing The first man is wearing a black jacket andlikelyars to) and has a serious expression. The second man is wearing glasses and a black sweater, and he appears to be listening intently. The details match with the description provided.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b1d11bbf05803067bca7d36eb26f9734c9cf0789
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ca8db5ee604eaeff96baecb6a55cb082a48fd65ded3d520920b959a5ec0db29
+size 74771
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..620864e44c810e6fafd47d45a30f22af46c5c4e9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman in a blue shirt standing in a kitchen. She is wearing a necklace and has her hair pulled back. The kitchen is equipped with a black stove and a white vase with flowers on the counter. The woman appears to be looking down at something, possibly a recipe or a cooking utensil. The overall style of the video suggests that it might be a cooking tutorial or a lifestyle vlog. The focus is on the woman and her actions, with the kitchen serving as a backdrop. The lighting is bright and even, highlighting the woman and the kitchen without creating harsh shadows. The colors are natural and realistic, with the blue of the woman's shirt contrasting nicely against the black stove and the white vase. The video seems to be well-lit and professionally shot, indicating that it might be intended for a larger audience, such as a YouTube channel or a cooking show."
+    ],
+    "video_ids": [
+        "JgDKjaEovLA_9_153to288"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a blue shirt, a necklace, her hair pulled back, a recipe or a cooking utensil, a white vase with flowers.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a blue shirt and a necklace, with her hair pulled back in a ponytail. She is holding a cooking utensil (a blender or food processor) and is in a kitchen setting. A white vase with flowers is visible on the countertop in the background. All elements described in the condition are present and consistent with the video content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a blue shirt, a necklace, with her hair pulled back. a ponytail. She is standing a green utensil,a green) similar processor) and there in a kitchen setting. There white vase with flowers is also on the counterop. front foreground. The elements match in the condition are present in match with the video content.\"\n}<|im_end|>\nGuidId",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d204cd118aa0e8e9edbbf33f841d2b7d2f8f1125
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec538fa77f3b9cdf3d1bd55359bc641799a55f2aea0f7bf920033f95d62a54c7
+size 112229
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..133a196cbd2fc69ecdea9b258d792357c1f97b7e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features three animated characters, each with a distinct color and design, standing on a round platform with a colorful, textured surface. The character on the left is a red, elongated creature with two antennae and a wide, open mouth, as if it's excited or surprised. The middle character is a yellow, elongated creature with a large, open mouth and a long, curled tongue, giving it a playful or mischievous appearance. The character on the right is a red, rounded creature with a smaller mouth and a more subdued expression. The background consists of a blue, textured wall with a large, jagged opening, and a gray, cylindrical pillar. The overall style of the video is cartoonish and whimsical, with vibrant colors and exaggerated features that give the characters a lively and engaging appearance."
+    ],
+    "video_ids": [
+        "T5r4ZKpHwhw_85_26to242"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three animated characters - a red, elongated creature with antennae and a wide mouth, a yellow, elongated creature with a large open mouth and a long, curled tongue, and a red, rounded creature with a smaller mouth. A blue, textured wall with a large, jagged opening and a gray, cylindrical pillar.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully features the three animated characters as described: a red elongated creature with antennae and wide mouth, a yellow elongated creature with a large open mouth and curled tongue, and a red rounded creature with a smaller mouth. Additionally, the background includes a blue textured wall with a jagged opening and a gray cylindrical pillar, matching the description. All elements are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features ful three three animated characters as described: a red,ated creature with antennae and a mouth, a yellow elongated creature with a large open mouth and a tongue, and a red rounded creature with a smaller mouth. The, the blue includes a blue, wall with a largeged opening and a gray cylindrical pillar, which the description.\"\n The elements are present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1032f224815008d9b337c0c8231d2bb64fff2ef9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1155-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7314c1c59535731b047d432275f6047077aacdfcc752784111831188c5707616
+size 155463
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..afe7ff0bff79771d8f0176acb4be68755e14782c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two men are sitting at a cluttered wooden table, engaged in a conversation. The table is filled with various items, including a coffee maker, a cup, and a bowl. One of the men is holding a remote control, suggesting they might be watching something on a television. The room has a cozy and casual atmosphere, with a window in the background and a poster on the wall. The men appear to be relaxed and comfortable in each other's company. The video captures a moment of everyday life, with the men enjoying their time together."
+    ],
+    "video_ids": [
+        "SWcfrQqfeRg_61_0to142"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, a coffee maker, a cup, a bowl, a remote control, a television (implied).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men sitting at a table. A black Keurig coffee maker is prominently featured in the center, and there are white bowls and cups on the table. While a remote control is not visibly present, a television screen is behind them displaying content, which satisfies the 'implied' condition. All core objects mentioned in the description are present or implied.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two men sitting at a table with One coffee coffeeurig coffee maker is on placed on front scene of and there is two coffee and a on the table, The a remote control and not visible present, the television is is implied the, a, which is the impliedimplied' condition. The the elements are in the description are present, implied in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..850fdc1d7d18929ec48cca69aaa096ff54c9e1cc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8e0ca1111732934fc51fe027460cf1b1118c7d5c0f51d373ce51293b9a6a68b2
+size 129983
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..6310b144b4c02a38206f967fc50cc719d2e70781
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a kitchen, preparing a meal. She is wearing a striped shirt and is standing in front of a counter. On the counter, there are various ingredients and utensils, including a bowl, a spoon, and a knife. The woman is pointing at the ingredients, indicating that she is in the process of cooking. The kitchen is well-equipped, with a refrigerator, an oven, and a sink visible in the background. The overall style of the video is casual and homey, capturing a typical day in the life of a home cook."
+    ],
+    "video_ids": [
+        "XFREBlRD1TM_0_0to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a counter, a bowl, a spoon, a knife, a refrigerator, an oven, and a sink.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman standing in a kitchen. A counter is visible with various items including a large metal bowl, a spoon (implied by the context of mixing), a knife (visible in a utensil holder), a refrigerator (behind the woman), an oven (on the right side of the frame), and a sink (on the far right). All specified objects are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a woman in in a kitchen, There counter is visible with various kitchen on a bowl bowl bowl, a green,partplied by the context of cooking), a knife (im on the holderil holder), a refrigerator,partind the woman), an oven (to the right side), the frame), and a sink (part the left left). The the objects are present and the with the scene.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4c8356de4bf9a9827bc6b0534e9a5c33f43245a7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee711b3d251b8558b0f487056b3d8a998fe013270e6c3e7c146e4f6a9ef5451f
+size 174249
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..7ba69f193e62831dd5a2f18d7dbec85aba30e888
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are sitting on a couch, engaged in a conversation. The man is wearing a black hat and a black shirt, while the woman is dressed in a denim jacket and a glittery top. They are both smiling and appear to be enjoying their conversation. The setting is a room with a large window, through which palm trees can be seen, suggesting a warm and sunny location. The overall style of the video is casual and relaxed, capturing a moment of friendly interaction between the two individuals."
+    ],
+    "video_ids": [
+        "7nAl-4lmezw_1_1014to1165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman sitting on a couch, which fulfills the 'Object(s)' condition. The man is wearing a black hat and shirt, and the woman is wearing a denim jacket over a sequined top. Their presence and appearance are consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two man and a woman sitting on a couch, engaging fulfills the 'Object(s)' condition. The individuals is wearing a black hat and a, while the woman is wearing a blue jacket. a blackined top. Both attire and interaction match consistent with the description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0856a2dbe8955abdc981053ed5cc799ff8139db5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:566b2932ce949b85ed6623a0c282ce776b8668c4b3b540f6e87a4bd1c9dd6f46
+size 164689
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..3cca5c2388f187d993e1e7ad6c9c3e7084ed42bc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a black shirt with a patterned design, sitting in a studio setting. He appears to be engaged in a conversation or interview, as suggested by his attentive gaze and thoughtful expression. The studio has a modern design with a blue background and a brick wall, which adds a touch of urban sophistication to the scene. The man's attire and the studio's design suggest a professional or formal context for the video. The overall style of the video is polished and well-lit, with a focus on the man's face and upper body, indicating that the content is likely centered around his thoughts or opinions."
+    ],
+    "video_ids": [
+        "8wEXIZ51lQE_43_0to150"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black shirt with a patterned design, sitting and engaged in a conversation or interview.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black shirt with a patterned design, sitting and actively engaged in conversation, which aligns with the described condition. The presence of a red chair and background elements does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black shirt with a patterned design, sitting in appearing engaged in what or which aligns with the description '. The background of a background brick and a elements does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..305565ffc1fd0a25d47c13d2610c8d11e9ad5005
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ad2128978e643a6d49c6f2d74eef562306e8edf1e944a92a41a65c339850b07f
+size 127219
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..56a078cd4bdaa83d367383301789908cbfffeb99
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a white plate with six cupcakes, each decorated with a swirl of blue and green frosting. The cupcakes are arranged in a circular pattern on the plate. The plate is placed on a blue tablecloth. The cupcakes are the main focus of the video, and the camera captures them from a top-down perspective. The video is likely a food-related video, showcasing the cupcakes' presentation and decoration. The style of the video is simple and straightforward, focusing on the cupcakes without any additional elements or distractions."
+    ],
+    "video_ids": [
+        "2WPnycuBdkg_0_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white plate with six cupcakes, each decorated with a swirl of blue and green frosting.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white plate with multiple cupcakes, each topped with a swirl of blue and green frosting, matching the description. Although the exact count of six is not explicitly confirmed, the visible cupcakes are consistent with the described decoration and setting.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a white plate with four cupcakes, each decorated with a swirl of blue and green frosting. which the description provided The the number number of cupcakes cupcakes not specified mentioned, the presence cupcakes align consistent with the description appearance and color.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5826006ba097520b6ccf8f014fee28dd99b6e295
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:417075535f521049cc36144d34926ae2f3df91075dc0c7515c6824d749873670
+size 110422
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..28193d3cae927b7a9896134c81e8e1cd9f340e25
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a leopard in a zoo enclosure, eating from a bowl. The leopard is spotted and has a long tail. The enclosure is fenced and has plants in the background. The bowl is white and contains red food. The leopard is eating from the bowl, which is placed on the ground. The video is a simple, straightforward depiction of a leopard in a zoo, with no additional action or movement. The focus is on the leopard and its feeding behavior."
+    ],
+    "video_ids": [
+        "pszvM49JCho_6_945to1110"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A leopard and a white bowl containing red food.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a leopard (or a similar big cat with spotted fur) eating from a white bowl that contains red food, which matches the 'Object(s)' condition. The presence of a fence and some plants in the background does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a leopard eatingthe a similar large cat) a fur) eating from a white bowl that contains red food, which matches the descriptionObject(s)' condition. The presence of additional cow and some green in the background does not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b205099c61982ca98e5ec18cac87e660710faa7f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d028cc59d2ce016539864e34f59fd60ab4f0f308d6911c1045afb2c929904b9
+size 266719
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..3e663a89f435efe31a481ed8eb679fd384c71a9c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a large number of uncooked dough balls, which are placed on a griddle or pan. The dough balls are round and appear to be made of a light-colored dough, possibly for making bread or pastries. The griddle or pan is filled with a large number of these dough balls, indicating that they are being prepared for cooking. The style of the video is a simple, straightforward food preparation video, focusing on the dough balls and the cooking process. The video does not contain any additional elements or background details, keeping the focus solely on the dough balls and the cooking process."
+    ],
+    "video_ids": [
+        "AM6wXynvzuA_7_341to467"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Uncooked dough balls, griddle or pan? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows numerous dough balls frying in oil within a pan, which matches the 'Object(s)' condition. The dough balls are visibly uncooked at the start but are being cooked in the pan, fulfilling the description. There are no conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a round balls that on a, a pan, which align the descriptionUn(s)' condition of The presence balls are un uncooked, the beginning and are being cooked, the pan, which the description of The are no additional elements in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7af00aa48d9efbb479d43dd55b1a87ddfc04c85f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3bb1f66198e3697939e3632164f06eb583ba6067e489c19e82e5348da4ecacb5
+size 122348
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..55985d4a5ca67546b9af5739f7d797b028b3665b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men engaged in a conversation. They are standing in front of a building with a red awning. Both men are wearing lanyards, suggesting they might be attending an event or conference. The man on the left is speaking, while the man on the right is listening attentively. The background is blurred, but it appears to be an urban setting with a car parked nearby. The overall style of the video is candid and informal, capturing a moment of interaction between the two men."
+    ],
+    "video_ids": [
+        "GPzZd351m3U_5_17to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men standing and conversing outdoors. Their presence and interaction are the central focus, and there are no elements contradicting the description of 'Two men'.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two men in and interactinging.. They attire and interaction fulfill the central focus, and there are no additional ining the description of 'Two men'.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..928c3c0768c250384cccfbeb82f42535e9fdc680
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1156-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:efceaba327867385ee4a06f81954d55bbd4d3ed476937877c0bba3e61d6f3b4b
+size 183498
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..039f349ce9ecf88ac53644644b792ded2233f716
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a panel discussion with a moderator and two guests. The moderator is a man with gray hair, wearing a white shirt and a microphone. He is seated in the center of the frame, facing the camera. To his left is a woman with long blonde hair, wearing a black dress. She is seated and facing the moderator. To his right is another woman with short blonde hair, wearing a floral print dress. She is seated and facing the moderator. The audience is visible in the background, seated in rows and facing the panel. The setting appears to be a television studio or a conference room. The style of the video is a standard panel discussion format, with the moderator asking questions and the guests responding."
+    ],
+    "video_ids": [
+        "kYtPezbPm4k_17_0to129"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with gray hair, a woman with long blonde hair, a woman with short blonde hair, and an audience.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with gray hair wearing a white shirt and a microphone, a woman with long blonde hair (partially visible on the left), and a woman with short blonde hair (in the middle background). The background also includes an audience, fulfilling all the described elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man with gray hair, a white shirt, dark woman, fl woman with long blonde hair inwially visible on the left), and a woman with short blonde hair (part a middle right). The audience includes includes an audience, which the the conditions elements.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5509e01c26ebc67c7487fba4ca417b1acbc4e1fb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:550a996b0ee9544d06a2018a5d7b32204cc3164d4c2b72390b48f27511d6b1bb
+size 134750
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..81248de8cd30473dc1e66ecd20a95eaa933b613d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a thrilling moment in a mixed martial arts (MMA) match. Two fighters, both shirtless and wearing gloves, are engaged in a fierce exchange. The first fighter, positioned on the left, is in the midst of delivering a powerful punch to his opponent. His muscles are tensed, and his focus is intense. The second fighter, on the right, is blocking the punch with his left arm, demonstrating his agility and skill. The background reveals a crowd of spectators, their attention riveted on the unfolding action. The atmosphere is electric, reflecting the high stakes of the match. The video is a dynamic snapshot of the intensity and athleticism inherent in the sport of MMA."
+    ],
+    "video_ids": [
+        "lyD2-nYmBYk_2_32to180"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two shirtless fighters, one delivering a punch, the other blocking it; a crowd of spectators.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two shirtless fighters in a cage, with one delivering a punch and the other blocking it. The background includes a crowd of spectators, which matches the described condition. The scene is consistent with a mixed martial arts fight, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts two shirtless fighters engaged a boxing, one one delivering a punch and the other blocking it. The background includes a crowd of spectators, which align the description scene. The video is consistent with a typical martial arts (, fulfilling the additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..41f432476003437c1e4488cc1913453468512e93
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:522902597f245fa7a1e062b16c99eceec63aa9fdda42c11c062f31235be845d5
+size 234517
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..1b5ad6c2c2455783879e69cdfbddb0fc3876debe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene dance practice session in a spacious studio. A dancer, dressed in a flowing pink gown, gracefully performs ballet movements. Her long hair is tied back, and she extends her arms elegantly, creating a harmonious blend of motion and stillness. The studio is minimalist, featuring white curtains on one side and a large mirror reflecting the dancer's movements. A small gray cat lies peacefully on the floor near the mirror, seemingly undisturbed by the activity around it. The room is brightly lit, enhancing the ethereal quality of the scene. The dancer's fluid motions are mirrored perfectly in the glass, adding depth to the visual experience. The overall atmosphere is calm and focused, highlighting the beauty of dance and the tranquility of the moment."
+    ],
+    "video_ids": [
+        "f7e50193b5723a8a3f92493eb43f59da3bff25780a90eb711df807947c26bf81"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A dancer in a pink gown, a small gray cat? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dancer in a pink gown performing gracefully in a dance studio, and a small gray cat is visible lying on the floor near the mirror. Both elements are clearly present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dancer in a pink gown performing ballet in a dance studio. which there small gray cat is present on on the floor near the dancer. The elements match present present and do the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e00bb569c77a70d17de198ae068d16e7396dd786
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9375e17edd5b1bfc2234fd5005dc56b0d1c1025aa4e8929db5c67ee966cc457d
+size 169748
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..7099ddb363b9e1d7c5805259aa627597084bd950
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red Honda car parked on a grassy field with trees in the background. The car is stationary and the focus is on the front of the vehicle. The style of the video is a straightforward, real-life depiction with no additional effects or animation. The car is the main subject of the video, and the surrounding environment provides context for the setting. The video likely captures the car's design and color, as well as the natural beauty of the outdoor location."
+    ],
+    "video_ids": [
+        "G4ehT5LPsv4_17_148to283"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red Honda car (front view)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red Honda car from a front view, matching the specified condition. The car is clearly visible, and the surrounding elements (grass, trees, sky) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a red Honda car from a front view, which the description '. The car is the visible, and the front environment,trees, flowers, and) do not contradict the description description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4da78d4c00ab875e5654efcdb8b5c071fb156c88
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:08f0ec44b4e602aad4ac485817c7b8aa27616ab19abf7ff880aebd106e5dd9e0
+size 173203
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..047a6d5be44ccfd10d718dc323adaa5e97ac32cc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a close-up view of a car's headlight in a dark setting. The headlight is illuminated, with the light shining brightly and casting a reflection on the car's surface. The car appears to be a luxury vehicle, with a sleek design and a shiny finish. The headlight is the main focus of the video, with the camera capturing the intricate details of the light's design and the reflections it creates. The video is likely a promotional or advertising video for the car, highlighting its design and features. The style of the video is sleek and modern, with a focus on the car's headlight and the reflections it creates."
+    ],
+    "video_ids": [
+        "9s-bCTpJ22g_30_46to218"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Car's headlight? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video focuses on a close-up of a car's headlight, showcasing its design and illuminated LED elements. The headlight is clearly the central object, and no elements contradict this focus. The surrounding car body and background are secondary and do not detract from the primary subject.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows on a car-up of a car's headlight, which its design and the state elements. The lightinglight is the the main object in and the other contradict the description. The lighting parts body and reflections are dark and do not detract from the primary subject,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1fa934f2ba057f74d51e08229ad80431e8c59100
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2dd315866a204019e363643ef7f6d9a91daf5a728f42af6a12cb7f421ec7118a
+size 167072
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..7426a9e03f4603c27128278c527a494a6ab9cb8d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing on a stage with a red circle beneath him. He is dressed in a blue shirt and a red tie, and he appears to be gesturing with his hands as he speaks. The stage is set against a black background with three large, illuminated circles. The style of the video suggests that it is a TEDx talk, which is a series of events that bring people together to share ideas and engage in conversation. The man's attire and the setting indicate that the event is formal and professional. The video captures the speaker's passion and enthusiasm for his topic, as well as the audience's interest and engagement."
+    ],
+    "video_ids": [
+        "6AtwNb4YjyM_9_16to153"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man standing on a stage with a red circle beneath him, dressed in a blue shirt and a red tie. There are also three large, illuminated circles behind him.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man standing on a stage with a red circle beneath his feet, wearing a blue shirt and red tie, which matches the core description. Behind him, there are three large, illuminated circular projections, also matching the description. The presence of the large red 'DX' letters on the left does not contradict the core description and is an additional element that is acceptable.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man standing on a stage with a red circle beneath him feet, wearing a blue shirt and a tie. which matches the description description. The him, there are three large, illuminated circles shapes, which align the description. The presence of the audience circles circleTED' logo on the floor side not contradict the core description and is likely additional element that does acceptable.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..02584086185156f18194f32dca0ce8ba02e672d3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bdf0c47b555e51359bcc793c92a1a4652471cfa0800bcd60ca0d273b23a068b5
+size 63085
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..df3816fe1dc19ce1f36bfac079e02944e7e80710
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a butterfly perched on a green leaf. The butterfly, with its vibrant orange and black wings, is the main focus of the video. It is seen in three different positions: facing the camera, facing away from the camera, and facing the camera again. The background is blurred, emphasizing the butterfly and the leaf it is perched on. The video is shot in a way that showcases the butterfly's beauty and the intricate details of its wings. The overall style of the video is a close-up nature shot, highlighting the butterfly's colors and movements."
+    ],
+    "video_ids": [
+        "2jwvdVUQGRQ_23_0to138"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Butterfly? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a butterfly with distinct orange, black, and white wing patterns resting on a green leaf, which matches the 'Object(s): Butterfly' condition. The butterfly's features and behavior are consistent with the description, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a butterfly with vibrant orange and black, and white patterns patterns, on a leaf leaf. which align the descriptionBut(s)' Butterfly' condition. The presence's wings, the are consistent with the description of and there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3d37cd4d27146e92cc6e5d8a206e7addef8f1d77
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fbef67eb816475c13eb1286564e9eb8822563a1dc5862e80357e15482dee71a8
+size 347889
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..1e0a15028ebe944669177158adbe69ea55a1c2f1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a car, focusing on the dashboard and steering wheel. The dashboard is black with red accents, and the speedometer and tachometer are prominently displayed. The speedometer shows a speed of 0 mph, while the tachometer indicates a speed of 1500 rpm. The fuel gauge is also visible, showing a full tank. The steering wheel is black and appears to be made of leather. The overall style of the video is a close-up shot, focusing on the details of the car's interior. The lighting is bright, highlighting the colors and textures of the dashboard and steering wheel. The video does not contain any text or additional objects."
+    ],
+    "video_ids": [
+        "u0Mqs_gq2og_32_24to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dashboard (black with red accents), speedometer (showing 0 mph), tachometer (indicating 1500 rpm), fuel gauge (full tank), steering wheel (black, leather-made).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a car dashboard with black background and red accents, a speedometer reading 0 mph, a tachometer indicating approximately 1500 rpm (around 1.5 on the scale), a fuel gauge at full (F), and a visible portion of a black steering wheel. All elements align with the description, with no contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dashboard dashboard with black and and red accents, a speedometer indicating 0 mph, a tachometer indicating  1500 rpm,though the145 on the scale), and fuel gauge showing full tank9), and a black black of a black leather wheel with The elements match with the description provided making no contradictions or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d6b5acfaa4bb55233b633ea5a8e64a9fbc508b92
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1157-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de5834801e0bba9f5a3d135c7be7adb7cd5e5f14f0a6f8f43ae35950bb7ceda4
+size 206159
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..d141a4e1eff968f6037496010b3685966ccb5774
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red BMW car being lifted by a car lift in a garage. The car is positioned on the lift, and the lift is in the process of lifting the car. The car is shiny and clean, indicating that it is well-maintained. The garage is well-lit, and there are various tools and equipment visible in the background. The style of the video is straightforward and documentary-like, focusing on the car and the lift without any additional embellishments or distractions. The video captures the process of lifting the car, providing a clear and detailed view of the car and the lift in action."
+    ],
+    "video_ids": [
+        "Lm_NDb97pYQ_27_0to147"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red BMW car and a car lift.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red BMW car elevated on a car lift in what appears to be an auto repair shop. The car's front is prominently displayed, and the lift mechanism is visible supporting the vehicle. The scene matches the described objects without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a red car car elevated on a car lift, a appears to be a automotive repair shop. The car is position is prominently displayed, and the lift is is visible, the vehicle. The setting align the description ' without any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..abb7efbcec596e337eed389e57573524696c6a85
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be230c09ec2270d9e0e0282dcf011ebec20d57609d65938cee96270ef67e1616
+size 124399
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..464f276fddf969c11c333c3311f6f1646db1a294
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the vibrant beauty of a tree adorned with bright yellow berries. The berries, round and plump, are scattered throughout the tree, creating a lush canopy of color. The tree is set against a backdrop of a blurred landscape, adding depth and context to the scene. The video is shot in a close-up style, focusing on the tree and its berries, while the background remains out of focus. This style emphasizes the tree and its berries, making them the central focus of the video. The colors in the video are bright and vivid, with the yellow of the berries standing out against the green of the tree and the blurred background. The overall effect is a visually striking representation of nature's bounty."
+    ],
+    "video_ids": [
+        "BWTRUcuUumE_6_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A tree with bright yellow berries? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a branch covered in bright yellow-orange berries, which aligns with the description of a tree with bright yellow berries. Although the branch is not part of a full tree, the berries are clearly visible and match the color description. The background elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a tree of with bright yellow berries berries, which aligns with the description of a tree with bright yellow berries. The the video is not a of a full tree, the focus and clearly visible and match the color and. The background is, not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0aac226e83a64af4d662b1aec050026a9c50bc63
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:df9b213e91fb0c93de4fb8c89e781a721c9e42b521bb719653d8cc3c77bb3c76
+size 183727
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..0039322f025e12cec5f15d06aa5f5719952b5e29
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a festive scene with a large bouquet of gold and black balloons. The balloons are arranged in a cluster, with the gold balloons reflecting light and the black balloons providing a stark contrast. The bouquet is held aloft, suggesting a celebratory event or a special occasion. The background features a chandelier, adding a touch of elegance to the scene. The video is likely to be a part of a celebration or a special event, with the balloons serving as a centerpiece. The overall style of the video is festive and celebratory, with the balloons and chandelier creating a visually appealing composition."
+    ],
+    "video_ids": [
+        "ZYI0vQNZ1Qk_24_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large bouquet of gold and black balloons arranged in a cluster.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large cluster of balloons, predominantly gold and black, arranged together as a bouquet. The balloons include various shapes such as numbers and round balloons, fitting the description. While other elements like chandeliers and walls are visible, they do not contradict the core description of the balloon arrangement.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large cluster of balloons in predominantly in and black in which in. a bouquet. The balloons are a shades and as spheres and stars ones, which the description of The the elements like aandeliers and furniture are visible, they do not contradict the main description of the balloons bouquet.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dc9b88be2a37ed8b3294f072dc4a5e2abd336877
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc4f343009b14a6781d35f41d18ea1d83c3783c05ac5e54ae0222a153056fc49
+size 132974
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..84e37b9c90f3b70b0ba3ca1558e180c16d9bdb17
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a brown and white guinea pig with curly fur. The guinea pig is sitting on a brown blanket with a blue polka dot pattern. The guinea pig is looking directly at the camera with its eyes wide open. The guinea pig appears to be calm and comfortable in its environment. The video is likely a close-up shot of the guinea pig, focusing on its fur and facial features. The style of the video is likely to be a simple, straightforward shot with no additional elements or actions."
+    ],
+    "video_ids": [
+        "EqTx9qzuSU0_18_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A brown and white guinea pig with curly fur.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a guinea pig with brown and white fur and curly hair, which matches the description. The guinea pig is resting in a cozy setting, and while there are additional elements like a blanket and a toy, they do not contradict the core description of the animal itself.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a closeinea pig with brown and white fur, curly fur, which matches the description provided The furinea pig's the on a colorful environment, and the the are other elements like a colorful with possibly colorful, they do not contradict the core description of the object.\"\n.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..eda56dee396041df74890b40019e671441a3392e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a907f322e8a9f8eefb1833522100d774ba925953d02a09837c0a4dff509dad3
+size 138720
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..add6b1ed41e02cf1acc19740447d28a8a9fec6b4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with glasses sitting on a purple couch. He is wearing a gray t-shirt and appears to be in a relaxed setting. The man is looking directly at the camera, suggesting that he is engaged in a conversation or an interview. The background is simple, with a white wall and a framed picture hanging on it. The overall style of the video is casual and informal, with a focus on the man and his surroundings."
+    ],
+    "video_ids": [
+        "1GJyOY9whPk_9_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with glasses, wearing a gray t-shirt, sitting on a purple couch and looking directly at the camera.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing glasses and a gray t-shirt, sitting on a purple couch, and looking directly at the camera. These elements match the core description provided. Additional elements, such as a framed picture on the wall and a light switch, do not contradict the description and are acceptable as background details.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing glasses and a gray t-shirt, sitting on a purple couch. and looking directly at the camera. The elements match the description description provided, There elements such such as the framed picture on the wall, the relaxed-colored, are not contradict the main and are acceptable.\"\n long details.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e724b6780bdd1e55e96e5fb59af2cbe561df8c82
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b6121544c46b2645d7c7789d15124963f50d4c25f9fca8cb07a83eaeadfb113
+size 170638
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..545b06094514035aabafd6d36dac588f56267c4c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a red and white race car with the number 16 prominently displayed on its side. The car is parked on a tiled floor, and the background is a clear blue sky with a few clouds. The car is equipped with a rear wing and a front splitter, indicating its aerodynamic design for high-speed racing. The car's body is adorned with various sponsor logos, including Motul, Kumho, and Zigstone. The overall style of the video is realistic, with attention to detail in the car's design and the surrounding environment. The car is the main focus of the video, and its position on the tiled floor suggests that it is either being showcased or prepared for a race. The clear blue sky in the background adds to the sense of anticipation and excitement that is often associated with racing events."
+    ],
+    "video_ids": [
+        "xP87MhbY-RM_2_0to135"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red and white race car with number 16, rear wing, front splitter, and sponsor logos (Motul, Kumho, Zigstone).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red and white race car with the number 16 prominently displayed on the side. It features a rear wing and a front splitter, as described. Sponsor logos such as 'Motul', 'Kumho', and 'Bridgestone' are visible on the car's body. The video matches the core description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red and white race car with the number 16 prominently displayed on the rear. The features a rear wing and a front splitter, which well. The logos for as MotMotul', 'Kumho', and 'Zridgestone' are visible on the car, body, The car does the description description of any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..63156b4cdca201b98a76f42d4121b020a5cec1e1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:878713ea4d2114c902a7ee7ff7b3db5331e0b80a21519c06d62863d0ada74d82
+size 186464
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..79ea6ed194c55dd40e086bbd7cadd6ed0f2322cf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a delicious meal being prepared and served. The first frame shows a white plate on a table, with a small pile of chopped green herbs and a silver fork. The second frame shows a white plate with a golden-brown egg on top of a toasted English muffin, topped with a slice of ham. The third frame shows the same plate with the egg and muffin, now covered in a creamy yellow sauce, and garnished with a sprinkle of green herbs. The style of the video is a close-up food preparation and presentation, with a focus on the textures and colors of the ingredients. The table setting is simple and clean, allowing the food to be the star of the show. The video is likely intended to showcase the dish's presentation and appeal to the viewer's appetite."
+    ],
+    "video_ids": [
+        "z9-WDNcnIJw_24_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: White plate, chopped green herbs, silver fork, golden-brown egg, toasted English muffin, slice of ham, creamy yellow sauce? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a white plate with a dish that includes a toasted English muffin, a slice of ham, a golden-brown poached egg, creamy yellow hollandaise sauce, and chopped green herbs (parsley) being sprinkled on top. A silver fork is visible in the background, which is consistent with the description. All core elements are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showcasing showing a white plate, a golden that includes a golden English muffin, a golden of ham, a golden-brown eggached egg, a yellow sauceaise sauce, and chopped green herbs.likelyley). being addedled on top. The silver fork is also in the scene, and is part with the description. The the elements are present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1c2ed4b74112ea1e4ccee0bb9049dad3bbd74c6f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c11aac88ca2eb421ea8b4a5f982539bf89b46d5708da9a22b7eca8f26bcb2121
+size 155393
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..e145b3a8b2e84d31d3ebf812bd9da8e7ce690b56
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a robot dressed in a white space suit with a gold helmet, standing in a space-themed environment. The robot is adorned with NASA and GM logos on its chest, indicating a partnership between the two organizations. The robot is positioned in front of a blue wall, which adds a sense of depth to the scene. The overall style of the video is futuristic and sleek, with a focus on the robot as the main subject. The lighting is bright and even, highlighting the details of the robot's suit and helmet. The video does not contain any text or additional objects, and the robot remains stationary throughout the frames. The video is likely intended to showcase the capabilities of the robot and the collaboration between NASA and GM."
+    ],
+    "video_ids": [
+        "llix751qov0_15_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A robot dressed in a white space suit with a gold helmet, adorned with NASA and GM logos.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a robot wearing a white space suit with a gold helmet, and both the NASA and GM logos are clearly visible on its chest. The description matches the visual content without contradiction, even though the background includes additional elements like machinery and equipment, which do not conflict with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a robot dressed a white space suit with a gold helmet, which it NASA NASA and GM logos are visible visible on the chest. The robot align the core elements of any.\"\n and though the robot is additional elements like a and a, which are not conflict with the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f869da8a58409c9e7951f42df06ec4d3e58a187a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1158-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:61693b65e05e1449a2f83ed7f867a5c19bdbc5e9f31ce05137fc7d5aaca6ebae
+size 187114
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..4beb832490358d4f9dcac5e8491549b10fac0db2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up, high-resolution, and detailed view of an engine bay, showcasing the intricate components and design of a vehicle's engine. The style of the video is informative and technical, focusing on the mechanical aspects of the engine. The engine bay is filled with various parts, including the engine cover, air intake, and other components. The video provides a comprehensive view of the engine's layout and design, allowing viewers to appreciate the complexity and engineering involved in creating a vehicle's power source. The video is likely intended for automotive enthusiasts or professionals, providing a detailed look at the inner workings of a vehicle's engine."
+    ],
+    "video_ids": [
+        "iqYcgWD71Ag_10_106to281"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Engine cover, air intake, other engine components? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the engine cover with the word 'CYLINDER' embossed on it, the air intake system, and various other engine components surrounding the cover. The camera pans out to reveal the entire engine bay, confirming the presence of all mentioned elements without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows an engine cover, a Renault 'RenLINDER' onossed on it, which air intake,, and various other engine components such the engine. The presence angle over to reveal more entire engine bay, confirming the presence of these the elements. any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a1a521b359f4917916a4883b0387ec0e7cec434b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f41e668969b24e8bd48a3757e05127f2f0be74952aa06adc927246e9f508694d
+size 176477
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..06d67e5b45a0e727d4e42795bfae7913d990fd47
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a vibrant underwater scene featuring a school of yellow and black striped fish swimming around a coral reef. The fish are darting and darting around the coral, their bodies shimmering in the light filtering through the water. The coral, a mix of orange and red, provides a stark contrast to the blue of the water. The fish are seen swimming in different directions, some closer to the camera, others further away, creating a sense of depth and movement. The video is a beautiful representation of marine life, showcasing the diversity and beauty of the underwater world."
+    ],
+    "video_ids": [
+        "G56T3AMvbG8_166_0to150"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A school of yellow and black striped fish? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a school of fish that are predominantly yellow with black stripes, matching the description. The fish are swimming around a coral structure, and while there are other elements like coral and possibly other fish, the main focus is on the school of yellow and black striped fish, which is accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a school of yellow that are predominantly yellow with black stripes, which the description of The fish are swimming in a coral reef, which the there are other colors like the and possibly other fish, the main focus is on the yellow of yellow and black striped fish, which ful the represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..29608e2eb05b0134f1134bb4cd49003d40bfa083
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:29a8e5b74c77d5ea2b82ea0a9b4265b2e984b2dcefc3fcc684acd6703c3d344e
+size 226736
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..b783b6bfaf2739a1cfa17ed7119f5f39f3538bbc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene yet dynamic scene featuring two tigers and a bird near a rocky water body. Initially, the two tigers are seen walking along the edge of the rocks, their orange fur with black stripes contrasting sharply against the dark gray stones. The larger tiger leads, followed closely by the smaller one, both moving cautiously towards the water. As they approach, a bird, possibly a heron or egret, is perched on the rocks, observing them. The bird remains stationary throughout the sequence, its white head and long neck standing out against the darker background.  The tigers continue their journey, carefully navigating the uneven terrain. The larger tiger occasionally looks back at the smaller one, ensuring it stays close. The smaller tiger follows attentively, its movements slightly more hesitant but determined. The water below reflects the surrounding environment, adding a sense of tranquility to the scene. The rocky backdrop provides a natural and rugged setting for this wildlife moment.  As the tigers reach the water's edge"
+    ],
+    "video_ids": [
+        "45bf97fcf438b76683f52e8079bd986fe1a2048a8c35222a35b8aa5e6e30f903"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two tigers and a bird? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows two tigers walking on rocks near water, and a bird (a heron) is visible in the foreground. Although there is a brief cut to a scene with deer, the primary focus of the video still fulfills the 'Object(s): Two tigers and a bird' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows ful two tigers and on rocks near a. which there bird isa whiteon) is also in the background. The the are a reflection moment to a different with a, the core focus remains the video is includesfills the 'Object(s)' Two tigers and a bird' condition as}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..76be0bd3f501bad650b3291e5676757ff51b82eb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a8671a32a9697e757c4b8dc3f93813b782c9a81ecbc297219a816db65408fa6
+size 305394
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..5789275d24d1a1144bb0cabc559ebd9e9c267874
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a gray beard and mustache, wearing glasses and a black jacket. He is seated in a room with a red wall and a window with blinds. The man appears to be speaking, as his mouth is open. The style of the video is a close-up shot, focusing on the man's face and upper body. The lighting in the room is soft, creating a warm atmosphere. The man's expression is serious, suggesting that he is engaged in a conversation or giving a speech. The overall tone of the video is professional and serious."
+    ],
+    "video_ids": [
+        "GKbqs3DlDIE_272_0to146"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a gray beard and mustache, wearing glasses and a black jacket.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a gray beard and mustache, wearing glasses and a black jacket, which matches the description. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a gray beard and mustache, wearing glasses and a black jacket. which matches the description provided The background and lighting elements are not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ad8c3ea71effcb66185cf53eee17e7aa0332705e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7cc86cb25ec68268a7edd0f330f302347f3544b930bf4d715f676de3ccedb696
+size 222051
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..50187a78f937aba3c7dfa835f0c6a6a2c41c7f17
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a vibrant and lush depiction of a tropical jungle scene. It features a waterfall cascading down a cliff face, surrounded by a variety of exotic plants and flowers. The waterfall is the central focus of the video, with its water flowing down the cliff and creating a misty effect. The plants and flowers are in full bloom, adding a burst of color to the scene. The video is shot in a way that captures the beauty and tranquility of the jungle, with a focus on the waterfall and the surrounding flora. The overall style of the video is realistic, with a high level of detail and a focus on the natural beauty of the scene."
+    ],
+    "video_ids": [
+        "2TdhmjbQS2c_14_0to104"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Waterfall, exotic plants, and flowers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a waterfall cascading down a rocky cliff, surrounded by lush greenery. Exotic plants, including large succulent-like flora and ferns, are prominently featured in the foreground and midground. While the video does not explicitly show flowers, the presence of colorful, small blooms near the base of the waterfall can be interpreted as flowers, fulfilling the condition. The scene is rich in natural elements consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a waterfall inading down a rocky cliff, surrounded by lush,ery and Theotic plants and including large leavesulent-like leaves, palms, are visible featured. the foreground and backgroundground. Vibr the plants is not show show flowers, the presence of vibrant bloss vibrant flowers in the base of the waterfall suggests be considered as flowers, fulfilling the ' of The overall is vibrant in detail elements that with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a99fafc5206b00dce7a716033f660e4c24c1b77b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1c369def670a841fa258428ddabcaa6cb8e64a128b853903406f5ea0cdf7207
+size 301079
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..0c022a6dc03d2467f0b6961890a48362597af3b6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a white jacket and blue shirt, sitting in a studio with a blue and white background. He is gesturing with his hands, indicating that he is in the middle of a conversation or explanation. The man appears to be the main subject of the video, and his actions suggest that he is the one speaking or presenting. The studio setting and the man's attire suggest that this could be a news segment or a talk show. The overall style of the video is professional and polished, with a focus on the man and his message."
+    ],
+    "video_ids": [
+        "19i72s3P9Kg_2_0to104"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a white jacket and blue shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a white jacket over a blue shirt, which matches the description. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing a white jacket and a blue shirt, which matches the description provided The man and the elements in not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a3ee112a38db468deaf00345d2a3dbce330ad503
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:df93e16278671600f0a1d065ffe0b90134dc78f16e153d93e3dd706bb67a0e66
+size 192038
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..36d8289c05cacf74f36540ec9c375fcdd5010d76
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a dog with a shaggy coat, sitting in a grassy field. The dog has a black and white face, with a pink tongue hanging out of its mouth. It is wearing a blue collar with a silver bell attached to it. The dog appears to be relaxed and content, enjoying the outdoors. The grass is lush and green, indicating that the video was likely taken during the warmer months. The dog's fur is long and shaggy, suggesting that it may be a breed with a thick coat. The overall style of the video is natural and candid, capturing a moment of the dog's life in its natural environment."
+    ],
+    "video_ids": [
+        "B9fPB2NimLE_25_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dog with shaggy, black and white coat, wearing a blue collar with a silver bell, having a relaxed and content expression with a pink tongue hanging out.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dog with a shaggy, black and white coat, wearing a blue collar with a silver bell, and it has a relaxed and content expression with its pink tongue hanging out. All elements of the description are accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a dog with a shaggy coat black and white coat, wearing a blue collar with a silver bell. and it has a relaxed and content expression with a pink tongue hanging out. The these of the description are present represented in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a9d2c96845d93e3e323ce007c0fdfb572f11685f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63f75f669eba4ba46fdaec10df937b95022448674345d65e53e94b80ab24bd41
+size 210732
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..9e9ffc5cb1f545ca0ed4f0b280ebcaa8c51aeab8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a white sports car parked on a street. The car is sleek and shiny, with a black license plate. The car is parked in front of a black tent with red and white lettering. The tent appears to be a booth or a tent for an event. The car is the main focus of the video, and it is shown from different angles. The car is stationary, and there are no people or other objects in the video. The style of the video is a simple, straightforward car show or car event video. The car is the star of the show, and the tent in the background adds context to the setting. The video does not contain any action or movement, and it is a still image of the car and the tent."
+    ],
+    "video_ids": [
+        "cEQHOhntdxI_40_0to159"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white sports car, a black tent? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully features a white sports car, prominently shown in multiple frames, and a black tent, visible in the background of one of the frames. These elements are clearly present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a white sports car and which displayed in the frames, and a black tent with which in the background with the of the frames. The elements match consistent present and match the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..77c618409017f424fe39fd4e60eb99cc01a66d84
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1159-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f2f2e640476f163309f0ee1ff9de796c9691f826b1e252a7cffe80847c950a3e
+size 147357
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..3267d463a09b4b8e1d9577452388af37c0591e1e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a first-person perspective from inside a car, showing the driver's view of the road ahead. The car's dashboard is visible, with a speedometer prominently displayed in the center. The driver's hands are on the steering wheel, which is black and has a textured grip. The car is moving along a road with trees on both sides, and the sky is visible through the windshield. The style of the video is realistic, capturing the experience of driving from the driver's perspective. The focus is on the car's interior and the road ahead, with no additional elements or distractions."
+    ],
+    "video_ids": [
+        "URLAedhIKbk_6_0to151"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Car dashboard, speedometer, steering wheel, trees, sky? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the car dashboard, speedometer, and steering wheel from the driver's perspective. Through the windshield, trees and sky are visible in the background, fulfilling all the specified objects in the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a interior dashboard, speedometer, steering steering wheel, the driver's perspective. The the windshield, trees and a are visible, the background, fulfilling the the conditions objects. the condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a6b7d2d5725bc237ca0ba26d8808af827adc56d5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8862c9877bedae6f6f62eeed19f3598c43bb815970015ff145841fdfc78f11bb
+size 266201
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..07b854ae1ee75bebec79ab480ab318ad27a1e8a6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a playful and colorful Easter-themed setup. At the center is a small, toy-like finish line sign with a checkered pattern, suggesting a race or competition theme. Surrounding the finish line are various chocolate eggs, some wrapped in shiny foil in vibrant colors like pink, gold, red, green, and blue, while others are plain brown. Among these eggs, there are two cracked brown eggs, each containing a small yellow rubber duckling, adding a whimsical touch to the scene. To the right, a large chocolate figure resembling a chicken is nestled among the eggs, enhancing the Easter motif.  The background consists of straw, giving the impression of a nest or a rustic setting. The text \"Best moments!!\" appears in the upper right corner, indicating that the video is meant to capture special or joyful moments. The camera remains stationary throughout the sequence, focusing on the detailed arrangement of the eggs, the rubber ducklings, and the chocolate chicken. The overall atmosphere is festive and"
+    ],
+    "video_ids": [
+        "3839aa3573126e6a7ccb6523b2ce75785b5da62d1136a51b47ab9f39507aea32"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small checkered finish line sign, chocolate eggs (some wrapped in foil, others plain), two cracked brown eggs with yellow rubber ducklings inside, and a large chocolate chicken figure.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a small checkered finish line sign with 'FINISH' written on it, multiple chocolate eggs (some wrapped in foil, others plain), two cracked brown eggs with yellow rubber ducklings inside, and a large chocolate chicken figure. All core elements described are present in the video without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a small checkered finish line sign, aFINISH' written on it, surrounded chocolate eggs (some wrapped in foil and others plain), and cracked brown eggs with yellow rubber ducklings inside, and a large chocolate chicken figure. The these elements of in present in the image, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2886dbe838a081acbccbc7d601fa3b8ead8d0da8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b15df43895182091c2aadc8c66862f481c9ff44ccc9a5d2f28c2687c5629f323
+size 74982
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..ca62f66098372a05fda96e608243ba5233ecccdb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a car driving along a winding mountain road. The car, a blue sedan, is seen from a high angle, giving a sense of its motion as it navigates the curves of the road. The road itself is a two-lane highway, curving gently around the mountainous terrain. The surrounding landscape is lush and green, with trees and rocks dotting the landscape. The sky above is clear and blue, suggesting a bright and sunny day. The overall style of the video is dynamic and adventurous, capturing the thrill of driving on a winding mountain road."
+    ],
+    "video_ids": [
+        "Ay3bCjG-xnk_26_0to138"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue sedan car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a blue sedan car driving along a winding road through a natural landscape. The car is clearly visible and matches the description of a blue sedan. There are no significant contradictions with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a blue sedan car driving on a winding road in a mountain landscape. The car is the visible and matches the description of a blue sedan. The are no additional additional or the given description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..aa15890c2120369aa12f7774647e0d7b9846451c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:43189466b2bbd4d7472666e2f77aa1ef2f3f22d4c3fd3219f89a97361c0130eb
+size 287673
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..3802caa3012bb21f4762e752a16fac030c0c1e48
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a formal press conference with four politicians standing behind a podium. The politicians are dressed in business attire, with two men wearing suits and ties, and two women wearing dresses. The podium has a logo on it, indicating the event is sponsored by Thomson Reuters. The audience is seated in front of the politicians, with some individuals taking photos or recording the event on their phones. The setting appears to be indoors, with a backdrop featuring the Canadian flag. The politicians are speaking to the audience, and the audience is attentively listening. The overall style of the video is professional and formal, capturing a significant political event."
+    ],
+    "video_ids": [
+        "3PLrkK-WFQM_4_28to164"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Four politicians, a podium, an audience, and Thomson Reuters logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows four individuals standing on a stage, which appear to be politicians, with one at a podium. An audience is visible in the foreground, and the backdrop prominently displays the Thomson Reuters logo. All core elements from the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts four individuals in behind a stage, which can to be politicians, behind a of a podium. There audience is visible in the foreground, and the Thomson includes displays the Thomson Reuters logo. The elements elements of the description are present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..242ccb9c93a18b99b91095be91b8bf15142a1bb8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b8631f1e84f4b323c8ac7d0f33aa05752f8c9e02ff0ae51a19c84d1df0a9207
+size 121217
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..5c799209aaf3d46f34fef9bfd8d785be47a6c099
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic scene of three sports cars racing down a winding road. The first car, a vibrant green Porsche GT3 RS, leads the pack, its sleek design and powerful engine evident. Following closely behind is an orange sports car, its bright color contrasting with the green of the first car. The third car, a red sports car, is slightly behind the orange car, adding to the sense of speed and competition. The road they're on is curvy and appears to be in a rural or less populated area, with trees and a power line visible in the background. The sky above is cloudy, suggesting an overcast day. The overall style of the video is dynamic and action-packed, capturing the thrill of high-speed racing."
+    ],
+    "video_ids": [
+        "rU8zy4071HM_13_0to119"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three sports cars: a vibrant green Porsche GT3 RS, an orange sports car, and a red sports car.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows three sports cars in a row: a vibrant green Porsche GT3 RS (identifiable by its model name and color), an orange sports car (likely a Lamborghini Gallardo, based on its design), and a red sports car (likely a Ferrari, based on its shape and color). All three vehicles match the description provided, and no elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows three sports cars: motion rural: a vibrant green Porsche GT3 RS,whichifiable by its design and and design), an orange sports car,which a Porscheorghini givenardo, based on its design and and a red sports car (which a Ferrari  based on its design and color). The three cars are the description of, and there additional contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1ba470f57579805e925b7a23338cf6e460f9530d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c0056789446deb1991e460d798706ecd1fadec812f9ba40d93b82d508026c862
+size 194149
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..e2f1793889d0a35b8796f5eafce64631116b282b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a blue tank top standing in a room with a door and a window. He is smiling and appears to be in a good mood. The room has a warm and inviting atmosphere. The man is the main subject of the video, and his cheerful demeanor is the focal point. The room provides a simple and uncluttered background that allows the viewer to focus on the man. The video is likely a casual, personal video, possibly taken during a social gathering or a casual day at home."
+    ],
+    "video_ids": [
+        "LJw5sesT2S4_15_0to110"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue tank top, smiling and in a good mood.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue tank top, and he is consistently smiling and appears to be in a good mood throughout the frames. His facial expressions and body language align with the description, and there are no conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue tank top, and he is smiling smiling, appears to be in a good mood. the frames. The posture expressions and body language suggest with the description of indicating there are no additional elements in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..faa13b2db67048dce5d5c8365f83f385d32a56be
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:945f2a6997908326424fcbc0558bdfc83a242bb4b27916c0981531b6d43f2715
+size 114037
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..e950e40a35eb137126ce13e92e9cf10c5ed3533d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a close-up view of a pot filled with a hearty soup. The soup is simmering gently, with steam rising from its surface, indicating it is hot and freshly cooked. The broth appears rich and golden, suggesting it might be a beef or chicken-based soup. Within the broth, there are visible chunks of meat, likely beef or chicken, which have been cooked until tender. Slices of onions are scattered throughout, adding a layer of texture and flavor. Small pieces of carrot are also present, providing a pop of color and hinting at the inclusion of vegetables in the dish. The overall scene suggests a comforting and nourishing meal being prepared, with the focus on the ingredients and the cooking process. There are no significant changes or movements in the video; it maintains a steady shot of the simmering soup, allowing viewers to appreciate the details of the dish."
+    ],
+    "video_ids": [
+        "3fadad70ee405c34ba58698bf2aa4436f71a95ae8bb2ea55f0de61a0a4dfa409"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Pot, hearty soup, steam, chunks of meat (likely beef or chicken), slices of onions, small pieces of carrot? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white pot containing a hearty, steaming soup with visible chunks of meat, slices of onions, and small pieces of carrot. The steam rising from the pot confirms it is hot and actively cooking, matching the description accurately.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a close bowl containing a hearty soup steaming soup. visible chunks of meat, likely of onions, and small pieces of carrot. The presence rising from the pot and the is hot, ste cooking, which the description of.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..58e6b01240e4f4ca703fabf617d3a653e042d07c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f26a06f2690e49874741611f6ea76c3979878a505178ec35fc4d5b0207cee1aa
+size 191412
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..f42f20d1ddf0d9d4bdbd3616bdc3d525674188ac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a thrilling snowboarding adventure on a snow-covered mountain. The snowboarder, dressed in a vibrant red jacket, is seen carving through the pristine white snow, leaving a trail of tracks behind. The mountain's rocky terrain is partially exposed, adding a rugged charm to the scene. The snowboarder's dynamic movements and the breathtaking mountain backdrop create a sense of exhilaration and freedom. The video is shot from a high angle, providing a bird's eye view of the snowboarder's daring descent. The clear blue sky above and the untouched snow below create a serene and beautiful contrast. This is a video that encapsulates the thrill and beauty of snowboarding in a stunning mountain landscape."
+    ],
+    "video_ids": [
+        "cgaY2hd7A-0_38_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A snowboarder in a vibrant red jacket.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a snowboarder wearing a vibrant red jacket descending a snowy mountain slope. The snowboarder is clearly visible against the white snow and rocky terrain, and their red jacket stands out, matching the description. The surrounding environment and the snowboarder's actions are consistent with the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a snowboarder wearing a vibrant red jacket, a snowy mountain slope. The jacketboarder is the visible and the white snow, the terrain, and the red jacket stands out, fulfilling the description of The presence environment, the actionboarder's posture are consistent with the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f66778a88861c417ec60b8327a8cee3cb2b64f9a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-116-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fef628d85fb88e0cb5deeff1c960db8facc6d5ae326c27c24c8209aa9388857e
+size 287700
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..cb2076a77edae78e2342b074bf1320b3666ed72f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman in a kitchen, preparing a green smoothie. She is smiling and holding a glass of green liquid, which is likely the smoothie she is making. The kitchen is well-equipped with a blender and various kitchen utensils. The woman is wearing a blue and white dress, which adds a pop of color to the scene. The overall style of the video is casual and homey, with a focus on the woman's enjoyment of the cooking process. The video likely includes shots of the woman blending the ingredients, pouring the smoothie into the glass, and possibly tasting it. The kitchen setting suggests that the video may be part of a cooking tutorial or a lifestyle vlog."
+    ],
+    "video_ids": [
+        "FqHh9tAsWHQ_9_101to253"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a glass of green liquid, a blender, various kitchen utensils.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman holding a glass of green liquid, standing next to a blender (juicer), and various kitchen utensils are visible on the counter. All core elements from the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a glass of green liquid, which in to a blender,whichicer), and various kitchen utensils and visible in the counter. The elements elements of the description are present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2ecf5d39d36e7dd0fae10ef4bd75dcfd7b3475d1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a08800cd8eca5ce8153ccf42d4d14b785f147e4873fe41cae52ae829f5ffc66
+size 271513
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..000af638db2f4a6b232585ad2f645b2117545484
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young man standing against a black background. He is wearing a blue long-sleeved shirt and has a beard. His expression is neutral, and he is looking directly at the camera. In the top left corner of the video, there is a logo consisting of a green circle with a white \"G\" inside. The style of the video is simple and straightforward, with no additional elements or distractions. The focus is solely on the man and the logo."
+    ],
+    "video_ids": [
+        "WqHUQMdXDQQ_0_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Young man, green circle with a white 'G'? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young man speaking directly to the camera, and a green circle with a white 'G' is clearly visible in the top left corner. These elements match the specified 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young man wearing, to the camera, which there green circle with a white 'G' is visible visible in the top left corner of The elements match the description 'Object(s)' condition without any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8f9b7affd5c7f8ecc4b5fbd1580b0f22ad293853
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:924088257d39daeefd421fa39455fd9202e2baa9071bbda98579dd2e0ed81d43
+size 68460
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..fd22c2d3195c8db44c28267d6124c1fd9c2704ad
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a zebra crossing a paved road in a savanna landscape. The zebra, with its distinctive black and white stripes, is seen walking from left to right across the frame. The road is bordered by tall grass on both sides, and the background features a mix of green and dry vegetation, with distant hills visible under a clear sky. The camera follows the zebra's movement smoothly, maintaining a consistent distance, allowing viewers to observe the animal's steady pace and the surrounding environment. The lighting suggests it is either early morning or late afternoon, casting soft shadows and enhancing the natural colors of the scene. The overall atmosphere is serene, highlighting the zebra's graceful motion against the backdrop of the savanna."
+    ],
+    "video_ids": [
+        "3619dad2284c8c548ea38e0ae19c21750d8eedfa68ecc0e28b33312ff3bada41"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A zebra? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a zebra walking along a road in a grassy savanna environment. The zebra is the central and only animal subject, and its appearance and movement are consistent with the description of 'A zebra'. No contradictory elements are present that would negate this condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a zebra crossing along a road, a naturaly landscapeanna landscape. The zebra is the central focus primary object in in and it presence and behavior are consistent with the description of aA zebra'. The other elements are present.\"\n would suggest the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..52d60846f739cdce01b86b9b35e3ef7731766dde
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8068668b17d8551e2aa616a6df1246e81953bf1472ce2b8055af8081610f444d
+size 220004
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..a4915697a5b0351b9840027c79c439f0b1e24d36
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a news anchor standing in front of a colorful background. The anchor is dressed in a dark suit and tie, and appears to be speaking into a microphone. The background consists of a large screen displaying various logos and images, including a cityscape with bright lights and a news ticker. The overall style of the video is professional and polished, typical of a news broadcast."
+    ],
+    "video_ids": [
+        "VRtfgc8AXXY_6_22to193"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A news anchor dressed in a dark suit and tie, standing in front of the background and speaking into a microphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed in a dark suit and tie, standing in front of a news studio background with illuminated cityscape graphics. He appears to be speaking, and while a microphone is not visibly attached to him, his posture and mouth movements suggest he is addressing an audience, consistent with a news anchor's role. The background and attire match the description, and no elements contradict the core requirement.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person dressed in a dark suit and tie, standing in front of a colorful studio background. a graphicsscape graphics. He appears to be speaking into which there the microphone is not directly attached to his, the posture and setting position suggest he is speaking an audience, which with the news anchor's role. The setting and setting align the description provided fulfilling there conflicting contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b6f971c00dd0c2ab7f93bfa300ba3dbe0dedaab0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cae51301ae0fcd03f84db2057ebaa6d74aab6794473d3416d86155d59264f3fd
+size 70126
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..3d5e009f66fd3e94581952725e62f9d47ba35055
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a dog statue in three different stages of being made. The first stage shows the dog statue in its raw, unpainted form, with visible clay or plaster texture. The second stage shows the dog statue with a base coat of paint applied, giving it a rough, unfinished appearance. The third stage shows the dog statue fully painted and detailed, with realistic fur and skin textures, and a lifelike expression. The style of the video is a time-lapse or stop-motion animation, showcasing the transformation of the dog statue from its initial state to its final, finished appearance."
+    ],
+    "video_ids": [
+        "5hdxQUAjo88_28_38to161"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dog statue? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video displays a highly detailed, realistic cake shaped like a bulldog lying down, which matches the description of a 'Dog statue' in terms of form and appearance. The cake's lifelike features, including fur texture, facial expression, and posture, convincingly represent a dog statue made from edible materials. Additional text about cake-making classes does not contradict the core object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts two series detailed and AI depiction sculpture like a dogog. down. which align the description of a 'Dog statue'. as terms of the and subject. The cake's textureelike features, such its texture and facial expressions, and posture, aligningly simulate a dog,. of a material. The elementsural the sculpt techniques and not contradict the core description description.\"\n}<|im_end|>\nGuidId",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3bb764c93089ccc9347b4b92f328c27e874a90a1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:431745f25f61feffd7287ed8263047e11eb4fb0951c26ccdd511c2b16d6dabe1
+size 286743
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..3576496c4524ec56d17fcb8129683a905a6cb8a5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a promotional advertisement for a company named Adventure Adrift. It features a scenic aerial view of a desert landscape with a winding road and a few scattered trees. The text overlay on the video reads, \"It's our Patrons that keep Adventure Adrift\" and \"Become a Patron today!\" The style of the video is a combination of aerial photography and text overlay, designed to encourage viewers to become patrons of the company. The overall tone of the video is positive and inviting, with a focus on the beauty of the desert landscape and the company's appreciation for its patrons."
+    ],
+    "video_ids": [
+        "IsiHN-BIa5M_4_0to152"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Desert landscape, winding road, scattered trees? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by prominently featuring a desert landscape with rolling hills, a winding road cutting through the terrain, and scattered trees dotting the hillsides. These elements are clearly visible and align with the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting featuring a desert landscape with a sand and a winding road that through the terrain, and scattered trees.ting the landscape.. The elements are clearly visible and align with the description provided.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6a2a411d18d9b64ffaf72ccca33ad7870f9cc494
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e2649e3988874c5f523ada944f6437b417fc21016be35936c21fc416c8946e0
+size 165843
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..f2633700d594a92bb98b85b8b1f9292b1575c4f9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a cooking tutorial, showcasing a variety of ingredients on a kitchen counter. The ingredients include chicken, salmon, black beans, chickpeas, strawberries, broccoli, carrots, blueberries, and bananas. The style of the video is informative and visually appealing, with a focus on the freshness and variety of the ingredients. The camera captures the ingredients in close-up shots, highlighting their textures and colors. The lighting is bright and even, ensuring that the ingredients are clearly visible. The video is likely to be accompanied by a voiceover or on-screen text providing instructions on how to prepare the ingredients and create a healthy and delicious meal."
+    ],
+    "video_ids": [
+        "xiARHAxc37s_3_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chicken, salmon, black beans, chickpeas, strawberries, broccoli, carrots, blueberries, bananas? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully displays all the specified items: chicken (in a white tray), salmon (on a white plate), black beans (in a can), chickpeas (in a can), strawberries (in a glass bowl), broccoli (whole head), carrots (three whole carrots), blueberries (in a glass bowl), and bananas (two bananas). All these objects are clearly visible and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video contains ful a the mentioned objects: chicken,represented the raw bowl), salmon,in a plate plate), black beans (in a bowl), chickpeas (in a bowl), strawberries,on a bowl bowl), broccoli (on and), carrots (s pieces and), blueberries (in a glass bowl), and bananas (two bananas). The these items are clearly visible and match the description provided}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7c6b13137c0cf11a9ea85d5fa5ade4bf8703269d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a38fe3de0d21f293673cea60643bcd4890d139ebf0e62ba3ec1ec26f82502e18
+size 116793
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..98112e3303f56aeb09c36a8982e389789785acb8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a bald man wearing glasses, sitting in front of a bookshelf. He is wearing a black shirt and appears to be engaged in a conversation or presentation. The bookshelf behind him is filled with various books, suggesting an intellectual or academic setting. The man's gestures and expressions indicate that he is actively participating in the discussion or explaining a concept. The overall style of the video is casual and informal, with a focus on the man's interaction with the viewer or his audience."
+    ],
+    "video_ids": [
+        "KMUmfl_KDuk_9_0to143"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bald man wearing glasses and a black shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man wearing glasses and a black shirt, which matches the core description. Additional elements like a bookshelf in the background do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a bald man wearing glasses and a black shirt, which matches the description description provided The elements such the bookshelf in the background do not contradict the main and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cb892ca408a6c51c5ac333c7be118428da833eea
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1160-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:91fbf4212aa4e6b9e7a5f6ec9ed4c76508718b8b780b17b7a1ee46ab05d99ecd
+size 163073
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..724afc38e12bc523a1d923ed2abcadd6d2412809
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two animated bears are interacting with a mailbox in a forest setting. The first bear is standing on its hind legs, holding a red apple in its paws, and appears to be placing the apple into the mailbox. The second bear is standing behind the first, holding an umbrella, and seems to be observing the action. The mailbox is attached to a wooden post and is located next to a small wooden cabin. The forest surrounding the scene is lush and green, with trees and bushes filling the background. The style of the video is cartoonish and whimsical, with a focus on the bears' actions and expressions."
+    ],
+    "video_ids": [
+        "oradtM46iRw_101_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two animated bears, a mailbox, a wooden post, and a small wooden cabin.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two animated bears, one holding a red object and the other holding an umbrella. A blue mailbox labeled 'MAIL' is visible near the wooden cabin, and there are wooden posts supporting the structure. The setting includes a small wooden cabin built into a tree, matching the core description. Additional elements like hanging lamps and bushes do not contradict the specified objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts two animated bears, a of a red apple ( the other standing a umbrella. There mailbox mailbox is 'POST' is visible on the bears post. which the is wooden posts in the cabin. The setting is a small wooden cabin, into the tree, which the description description. The elements like the plants and a are not contradict the core objects.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..413647be005b5033d5a00779f3650884e4ff8770
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f8e6b03ef9eb9ca1d6b18f8a31aacc502620efc1bcbf239bff5b457f96f6cd4
+size 185349
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..2315f524be1d1f3a25c207212fca1f64de428fc5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a bald man with a beard, wearing a plaid shirt, sitting in an office chair. He is looking to the side with a slight smile on his face. The office setting includes a potted plant in the background, and the lighting is bright and natural. The man appears to be in a relaxed and comfortable state, possibly engaged in a conversation or reflecting on a thought. The overall style of the video is casual and candid, capturing a moment of everyday life in a professional setting."
+    ],
+    "video_ids": [
+        "5y7H1vGgQu0_14_0to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bald man with a beard, wearing a plaid shirt, sitting in an office chair with a slight smile and looking to the side.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man with a beard wearing a plaid shirt, sitting in what appears to be an office chair. He is looking to the side with a slight smile, matching the core description. Additional elements like a laptop and background plants do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man with a beard, a plaid shirt, sitting in an appears to be an office chair. He has looking to the side with a slight smile, which the description description provided The elements such the plant and a plants are not contradict the main and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..037c926abb30a63cbaee4b652bc862c3e1a6f659
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a09cbe81c6f51ba9a8a435538311a50adc42b46f3530b0eda7719592b3679f1d
+size 291243
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..26d9eb33fbc52b3475fc6831d4f919645fff3d45
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person holding a bowl of enchiladas, which are a type of Mexican dish. The enchiladas are covered in a sauce and topped with a green garnish. The person is holding the bowl over a table, which has other dishes and utensils on it. The video also includes text that provides the price of the enchiladas in pesos and US dollars, as well as the price of a lemonade. The style of the video is casual and seems to be taken in a restaurant or a home setting. The focus is on the food and the price, rather than the surroundings or the person holding the bowl."
+    ],
+    "video_ids": [
+        "AOF7hqoZD90_28_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a bowl of enchiladas, a table, other dishes and utensils? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person holding a bowl of enchiladas, placed on a table with other dishes and utensils visible in the background. The core elements described are all present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows depicts a bowl sitting a bowl of whatiladas, which on a table. other dishes and utensils visible in the background. The presence elements of in present present and accurately with the description content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b06d9a98d39f27edd4dafda9e1988b04059007a4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1e9eb0239af5b2435c240be2856de745601a2d303e629f012cb459e1d1e7999
+size 98370
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..3bd77cbe634c08073b2b71be300a2f5468094332
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a tan luxury car, focusing on the driver's seat and dashboard. The car's interior is well-maintained and features a leather upholstery. The dashboard is equipped with a digital display and various controls. The steering wheel is also made of leather and has multiple buttons on it. The car's interior is clean and well-organized, with no visible clutter or damage. The car appears to be parked in a garage or a similar indoor setting. The video is likely a promotional or sales video for the car, showcasing its interior features and design."
+    ],
+    "video_ids": [
+        "HtnDw-q5CFw_28_0to188"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Driver's seat, dashboard, steering wheel? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the driver's seat, dashboard, and steering wheel as requested. The camera pans across the interior, providing a comprehensive view of these key elements, and no conflicting objects are present that contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows the interior's seat, dashboard, and steering wheel, described. The image angle across the interior of focusing a view view of these elements elements, which there additional elements are present.\"\n would the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..64cb0b8d78f063711e331feadbcd2e709adaea69
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:99e9169c73ad5dd296435daf49ed1d5151bf28eb32df4f6946d6489a6b151d70
+size 74190
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..2bb36b2032c97f05dba180879511af27781ed4d5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a steering wheel in a car, focusing on the center console and dashboard. The steering wheel is brown with a yellow Ferrari logo in the center. The dashboard features a speedometer and other gauges, and there are air vents on the right side. The steering wheel has several buttons and a red button on the left side. The style of the video is a straightforward, static shot with no movement or action, focusing on the details of the steering wheel and dashboard. The lighting is bright, and the colors are vibrant, highlighting the brown and yellow of the steering wheel and the black and white of the dashboard."
+    ],
+    "video_ids": [
+        "qJKOhmc-EMg_14_0to143"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, center console, dashboard, speedometer, gauges, air vents, buttons, red button? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing the steering wheel with the Ferrari logo, center console, dashboard, speedometer, gauges, air vents, buttons, and a red button (engine start). All specified elements are visible and accurately represented in the frame.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showcasing showing a steering wheel, a Ferrari logo, the console, dashboard, speedometer, gauges, air vents, and, and a red button.likely start/ The these objects are present and match represented in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..535c709431eedc873682296ce5a82fe8b8fd91ac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7188bd5aa937d2ea02cd3ff0ed99112f01169f5cd1e8b682318e921beaa5687f
+size 156145
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..c19e446e5135187cee93bf24115a093e5abdf89d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a breathtaking aerial view of a mountainous landscape, showcasing the stunning beauty of nature and the serene tranquility of a lake. The first frame captures the majestic mountains, their peaks reaching towards the clear blue sky, while the second frame reveals the calm lake nestled among the mountains, its surface reflecting the surrounding scenery. The third frame provides a closer look at the lake, revealing the lush greenery of the surrounding landscape and the quaint town nestled on its shores. The video is a testament to the awe-inspiring beauty of nature, captured from a bird's eye view."
+    ],
+    "video_ids": [
+        "bDCTVFOtKvw_112_24to168"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ['mountains', 'lake', 'lush greenery', 'quaint town']? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing mountains surrounding a large lake, with lush greenery covering the hillsides and a quaint town nestled along the shoreline. All specified elements are prominently and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting mountains, a lake lake, with lush greenery visible the slopesides. a quaint town situated along the shore. The the elements are present featured co depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..74e697f3dc6a6f456c65610edfff63f3e3f7687d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0c130187414af2257d42329141ab0d39c75f20a6adc6bccd0ec259795275bab
+size 168437
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..afaf3b94fa7635b4c8429468427ab0fa7d86631f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a black and white time-lapse of a man driving a car. The man is wearing a suit and tie, and he is seen in various positions as he drives. The car is a luxury vehicle with a steering wheel and dashboard visible. The man's hands are on the steering wheel, and he appears to be focused on the road. The car is moving, and the man's position changes as he navigates the vehicle. The time-lapse effect creates a sense of motion and progression. The black and white color scheme adds a timeless quality to the video."
+    ],
+    "video_ids": [
+        "adh6s4AKRVg_12_47to198"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man, car (luxury vehicle).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a man seated inside a luxury car, specifically a Bentley, which matches the description. The man is visible in the driver's seat, gesturing as if speaking, and the interior of the car, including the steering wheel, dashboard, and seats, is prominently displayed, confirming it is a luxury vehicle.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as depicting depicting a man driving in a car vehicle. driving a sedan, as is the description of The man is dressed wearing a driver's seat, anduring as if he, and the car of the car, including the steering wheel and dashboard, and door, is visible displayed. indicating the is a luxury vehicle.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..34ec473499fffd3460f5eec92dde457059fbc5c6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e8773d0b6da62cb7b72b91c06968c9d5c12c19a1eef42f236cbde6e86ca3591
+size 232679
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..2888beb5eac2509755fbeacda96337f63a776315
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen interacting with a large alligator in a controlled environment. The man, wearing a white baseball cap and a gray t-shirt, is kneeling down next to the alligator, which is lying on its back on the ground. He appears to be examining or treating the alligator, possibly as part of a wildlife conservation or rehabilitation effort.  In the background, there are other people observing the scene, suggesting that this might be a public demonstration or educational event. The setting appears to be outdoors, with trees and a fence visible in the background. The overall style of the video is informative and educational, aiming to raise awareness about alligator conservation and the work being done to protect these animals."
+    ],
+    "video_ids": [
+        "FHqDjGV6YL8_3_0to181"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a large alligator, and other people observing the scene.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting on a large alligator while other people are present around them, observing or assisting. The core elements \u2014 a man, a large alligator, and other people \u2014 are clearly visible and consistent with the description. Additional elements like clothing or background details do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man interacting next the large alligator, other people are observing in the, observing the standing. The scene elements of a man, a large alligator, and other people \u2014 are all visible and match with the description. The elements like the and the details do not contradict the main description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b65f163865cfaa16ae4d35f8e1f610debc3f2929
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1161-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21e570876f9a3c3e82c2c5755eb555797ed3e3f454c26bb76cef2a924751e2d1
+size 227895
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..ebee9f2f226c9683caf4e76cbb74351309cc1a76
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young girl and a boy are seen interacting with a tray of cupcakes. The girl, wearing a striped shirt, is seen placing a cupcake into a wrapper. The boy, standing behind her, watches her actions. The tray of cupcakes is placed on a wooden table. The scene takes place in a room with a brick wall in the background. The video captures a moment of creativity and learning, as the children engage with the cupcakes."
+    ],
+    "video_ids": [
+        "_R4lsHnh8qc_15_199to320"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young girl, a boy, and a tray of cupcakes.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a young girl and a boy interacting with a tray of cupcakes. The girl is seated at a table, holding and gesturing toward the tray, while the boy approaches and also interacts with the tray. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a young girl and a boy standing with a tray of cupcakes. The girl is seen and a table, and a decoratinguring towards the cupcakes, while the boy is her looks interacts with the cupcakes. The elements elements of in present: any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fc54cb5269bb4b27ce2a0a9fd18c5e2b8036396d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5bdeacf6219dc4cb288a874c2475a25627e2920da13283351156248fdd8f8b8c
+size 185289
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..39ba80882c811c25fa1bf30f6953f3c6b10b268c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man with glasses is seen arranging white flowers into a clear, elegant glass vase. The vase is placed on a table, and the man is carefully placing the flowers, ensuring they are evenly distributed. The background features a green wall, adding a touch of color to the scene. The man's focus and attention to detail suggest that he is creating a beautiful floral arrangement. The overall style of the video is simple and elegant, with a focus on the man's actions and the beauty of the flowers."
+    ],
+    "video_ids": [
+        "PL_XKuEpeHw_19_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with glasses, white flowers, a clear, elegant glass vase, a table, green wall? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with glasses arranging white flowers in a clear, elegant glass vase. The setting includes a table and a green wall in the background, matching the described elements. Additional decorative items in the background do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man with glasses arranging white flowers in a clear, elegant glass vase on The setting includes a table and a green wall, the background, which the description elements. The elements elements like the video, not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..38e48ede28daddb62a4b724260b97c6315f26257
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:991cbb0d88d54462a295ada3cc47c69248fc7aba229f937ed6f23734d1b48b18
+size 197874
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..0536fb86a9cedebc8500875c7bd8956a0506b7cc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a car engine being worked on. The engine is open and various parts are visible, including the battery, alternator, and other components. The engine is covered in a protective cover with the Toyota logo on it. The car is parked in a garage or similar setting. The style of the video is a straightforward, unembellished documentation of the process of working on a car engine. The focus is on the engine and the tools being used, rather than any artistic or aesthetic elements. The video is likely intended for educational or instructional purposes, showing viewers how to work on a car engine."
+    ],
+    "video_ids": [
+        "9eQDQlcXSlY_16_0to129"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A car engine, battery, alternator, other components, and tools being used.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the engine bay of a car with visible components including a battery, alternator (or similar electrical components), and other engine parts. The camera pans across the engine bay, highlighting these elements, which aligns with the 'Object(s)' condition. Although no tools are visibly being used, the presence of the engine, battery, and other components fulfills the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts depicts a hood compartment of a car, various components such the battery and alternator,ind similar components components), and other engine parts. The presence angle over the engine bay, highlighting these components. which aligns with the 'Object(s)' condition. The the tools are actively being used in the presence of the engine components battery, and altern components isfills the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9cecc22f7d0f9c1421393b47adad73f74feaff4c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf14d6dbad8c6abff465bef65dc35b9ff0388af1e03d4fc7b31f628f92d9df71
+size 152952
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..dff19f27c9d923200e9e0117d0c0deadb5fbb34f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of disappointment and reflection for a football player. The player, wearing a white and blue uniform with the number 12, stands on the field with his head down, hands clasped together in front of him. The background is filled with other players, some of whom are wearing helmets, indicating that the game is still in progress. The player's posture and expression suggest that he is contemplating his performance or the outcome of the game. The overall style of the video is realistic, capturing the intensity and emotion of the sport."
+    ],
+    "video_ids": [
+        "e69MPU8jnf0_27_0to146"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A football player in a white and blue uniform with the number 12, standing with his head down and hands clasped together.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a white and blue uniform with the number 12, standing with his head down and hands clasped together. The player's posture and attire match the description, and while there are other players and background elements, they do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a white and blue uniform with the number 12, standing on his head down and hands clasped together. The player's posture and uniform match the description provided and there there are other players in a elements, they do not contradict the main description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..84550a85aa5a889cf362c4cfb7bc3d403ec39654
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:791e2d43dfe331403612767c7d9a0e7d513b203d6d306019596eeaa2b17ccda2
+size 224801
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..f15596204d88001ae2956c59d8f46c3fd6ab391b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman posing against a gray wall. She is wearing a black hoodie and pants, with her hood up. Her hands are placed on her head, and she is looking down. The style of the video is casual and relaxed, with a focus on the woman's outfit and pose. The gray wall provides a neutral background that allows the viewer to focus on the woman and her attire. The overall mood of the video is laid-back and comfortable."
+    ],
+    "video_ids": [
+        "HiHsnQt4uHg_9_24to155"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Woman posing with black hoodie and pants, hood up, hands on head, looking down? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a black hoodie and pants, with the hood up, her hands on her head, and she is looking down. These elements match the description exactly, and there are no conflicting elements in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a black hoodie and pants, with the hood up, and hands on her head, and she is looking down. The elements match the description provided, indicating there are no additional elements in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7cee347e51d7aeaf68dee1e23a2fb542ff9ed244
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d91af8c87d70f328e363f03dd40313dba11a2fce7bbb8ae86be60b64a702ff1d
+size 112781
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..3d39cdcec4d488cab05eb73679e2b4b769c0be35
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a news anchor named Mike Montecalvo, who is dressed in a suit and tie, presenting news on a television screen. The background of the video shows a large, white building with a dome, which is likely a significant landmark or government building. The style of the video is professional and polished, typical of a news broadcast. The focus is on the news anchor, who is delivering the news with confidence and authority. The overall impression is that of a serious and informative news report."
+    ],
+    "video_ids": [
+        "X9-T-pVy7Ag_0_0to174"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A news anchor named Mike Montecalvo dressed in a suit and tie, presenting news on a television screen.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man identified as Mike Montecalvo, dressed in a suit and tie, seated in front of a backdrop that resembles a news studio. The lower third graphic confirms his name and the channel's slogan, and the setting is consistent with a news broadcast. All elements align with the description of a news anchor presenting news on a television screen.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a person dressed as Mike Montecalvo, dressed in a suit and tie, standing at front of a news that includes a news studio. The presence third graphic displays the name, the setting's branding, which the setting is consistent with a news broadcast. The elements align with the description of a news anchor presenting news on a television screen.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a21f3fb20b5ba528e3a088a64da207a944b01fe5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:484fdbbc86bfd26ad49bc0a5ec14abd4f9ed03c5b8ddb6046da1d79f71d959d6
+size 154945
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..48f99065b716b034c30b01a2d72c680f1b49d15f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a dynamic scene of two airplanes in flight. The first airplane, a large one with a yellow and black checkered pattern on its tail, is flying in the foreground. It appears to be a vintage aircraft, possibly a fighter plane, given its design and the smoke trails it leaves behind. The second airplane, a smaller one, is flying in the background, following the larger plane. The sky is clear and blue, providing a stark contrast to the airplanes. The style of the video is realistic, with attention to detail in the aircraft and the surrounding environment. The video captures the motion and speed of the airplanes, as well as the sense of adventure and excitement associated with air travel."
+    ],
+    "video_ids": [
+        "k15BSV4Xx9w_79_0to128"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two airplanes, a large yellow and black checkered tail vintage aircraft and a smaller one.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two airplanes: a larger vintage aircraft with a yellow and black checkered tail, and a smaller aircraft in the distance. The description of the two planes matches the visual content, even though additional elements like smoke trails and explosions are present, which do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts two airplanes in a larger one aircraft with a yellow and black checkered tail and and a smaller one in the background. The larger accurately the ' airplanes and the video elements of with though the elements like the trails are a are present, which are not contradict the core description of}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4e594d1ca3c52721aff1280cc087ab2ec17de35d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b9264e9e28fdf8cf338e8ce8688d657f85eee4f783bbc60f4ea61139e10ad8e9
+size 207582
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..1b62df58c8abd74d710abace2cf2b67a6c03af08
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a black and white time-lapse of a cat's eyes glowing in the dark. The cat's eyes are the only colored elements in the video, with a striking green hue. The cat is lying down on a white surface, and its eyes are wide open, looking directly at the camera. The cat's fur is a mix of gray and white, and it appears to be a tabby cat. The background is dark, which contrasts with the cat's eyes and the white surface it is lying on. The video is shot in a close-up style, focusing on the cat's face and eyes. The time-lapse effect creates a sense of movement and change in the cat's eyes, making the video visually interesting and engaging."
+    ],
+    "video_ids": [
+        "6btQUn_NAcI_83_0to135"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Cat's eyes (green), cat's fur (gray and white), cat's body (lying down)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a cat with striking green eyes, gray and white fur, and a body lying down. These elements are clearly visible and match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close with green green eyes, which and white fur, and the body that down. The features match consistent visible and match the description provided.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2d21110175399cc090d2b738cda9850b84804a5c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1162-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4fd4aab83c0eec145175c6004adafb127eed6140ba30aa8eed24e49fcbae803a
+size 339841
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..836bb9615f3736744a5dbc3e9cd15459f2e60bb2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment on a football field. A player, donned in a white jersey with the number 39, is in the midst of a celebratory dance. His black helmet gleams under the stadium lights, and he's holding a football in his right hand, ready to make a play. The player's joy is palpable as he dances, his left hand raised in a triumphant gesture. The background is a blur of blue and white, indicating the presence of a crowd and the excitement of the game. The player's movements and the vibrant colors create a sense of energy and excitement, encapsulating the thrill of the sport."
+    ],
+    "video_ids": [
+        "vWN0Qbxxk1M_20_0to130"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a white jersey with number 39, a football, a black helmet.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player wearing a white jersey with the number 39, a black helmet, and appears to be in a football setting. Although a football is not visibly held or in motion in the frame, the player's attire and context strongly suggest a football scenario, fulfilling the core description. Additional elements like background personnel do not contradict the main object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a player in a white jersey with the number 39, a black helmet, and holding to be holding the football uniform. The the football is visible explicitly held by in the, the frames, the player's posture and the strongly suggest the football game. fulfilling the ' description of The elements like the details and not contradict the main description condition.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d395305a562635300fa1fc70c58d28b48af07c8d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:20121c3692eb8279df3d370963e51d5750ec8979abf514b96598a587e9fef3f5
+size 264366
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..06a4f458cfc16280c5251386f7614a4c782b4fce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a vintage car engine, focusing on the intricate details of the engine's components. The engine is silver and features a prominent air scoop on top. The video captures the engine from different angles, showcasing the craftsmanship and design of the vintage car. The style of the video is informative and detailed, providing a comprehensive view of the engine's components and their arrangement. The video is likely intended for car enthusiasts or those interested in the mechanics of vintage cars."
+    ],
+    "video_ids": [
+        "AbYvsPDfcv4_5_324to454"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Vintage car engine, engine components? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a vintage car engine with visible engine components such as carburetors, hoses, and metal parts. The engine is detailed and appears to be from a classic car, matching the description. The presence of a logo in the corner does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a vintage car engine with various engine components such as theuretors, hoses, and other parts. The style appears detailed and appears to be from an classic car, which the description of There presence of additional black or the corner does not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c5153fec4c74ce5cad3ca9f2e738f108c0b5e77a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e41d8f4d4137f73d171cabc817c3f218f2d32a9e0f97bf41471abe955c337acb
+size 256248
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..3020a6e1b793bb63145f6c51caa4c969084b917b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a documentary-style piece featuring a man in a factory setting. The man, who is the main subject, is seen in a close-up shot, wearing a green polo shirt. He appears to be in deep thought or concentration, as he looks off to the side. The factory in the background is filled with various machines and equipment, including sewing machines and a large mirror. The lighting in the factory is bright, illuminating the man and the machines. The overall style of the video is realistic and informative, providing a glimpse into the daily life of a worker in a manufacturing environment."
+    ],
+    "video_ids": [
+        "8g3MuraCw2k_9_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a green polo shirt, several machines and equipment, a large mirror.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a green polo shirt as the main subject, with several machines and equipment visible in the background, consistent with a factory or workshop setting. A large mirror is also visible in the background, reflecting part of the scene. These elements align with the 'Object(s)' condition described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing a green polo shirt, the main subject. and several machines and equipment visible in the background. which with the factory or workshop setting. There large mirror is also present in the background, fulfilling the of the scene. The elements align with the descriptionObject(s)' condition provided.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4c3d5c5f87decaddf1092c59ac97a36a91947ca2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:661bc9e804c3bce8975049776d42f0e402b5fa1b075061ca21285a9140582189
+size 98156
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..416e951c267cc602eb53df2e1f9096d8cb938768
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a large, prehistoric sea creature, possibly a plesiosaur, swimming through the ocean. The creature's long neck and head are visible above the water's surface, while its body remains submerged. The creature's skin appears to be a mix of gray and yellow, with darker stripes running along its neck and back. The ocean around the creature is a deep blue, with white foam and bubbles indicating the movement of the water. The background features a cloudy sky with rays of sunlight piercing through, creating a dramatic and somewhat ominous atmosphere. The style of the video is realistic, with attention to detail in the creature's design and the ocean's environment. The video likely uses computer-generated imagery (CGI) to create the scene."
+    ],
+    "video_ids": [
+        "ApBy9vU6JUA_16_66to223"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large, prehistoric sea creature (possibly a plesiosaur)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a large, long-necked, aquatic creature with a serpentine body and spines along its back, partially submerged in water, which strongly resembles a plesiosaur. The creature's appearance and environment align with the description of a prehistoric sea creature, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a large, pre-necked creature pre creature that a bodypentine body, flik, its back, which submerged in water. which align resembles a plesiosaur. The creature's appearance, the align with the description of a prehistoric sea creature, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8dda91bf139afc339bf09db34d78cdb1be9095de
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0842baad544cd7d781fd2b0e943f76ace45df570b229491e0344672a54268001
+size 296546
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..5a2d08b8160004615dad94ee56c65ee071a891de
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The image is a screenshot from a video game, featuring three characters in a forest setting. The characters are dressed in fantasy armor, with one character holding a sword. The text overlay is in Japanese, indicating that the game is likely in Japanese. The characters are standing in front of a forest, with trees and foliage visible in the background. The overall style of the image is reminiscent of a fantasy adventure game, with detailed character designs and a lush, natural environment."
+    ],
+    "video_ids": [
+        "0xwYjMgNAso_16_293to480"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three characters in fantasy armor, one holding a sword.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three characters in fantasy armor. The character on the right is holding a sword, and all three are dressed in detailed, fantasy-style armor consistent with the description. The background and character designs align with a fantasy setting, fulfilling the core condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts three characters dressed fantasy armor, Two character on the left is holding a sword, which the three characters standing in elaborate, elaborate-style armor. with the '. The presence and the design also with a fantasy theme, fulfilling the ' description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5fdc341d4293d76c01c8863ef576b3df92a28bc5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:097ddcb5ba8d31da8c83b50b1f0fb94ea3e48e02c49ae179d28f63f11f43ef73
+size 282732
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..adc5997eac7bc83c93ee3db8b7a0e6cc9233f754
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a plate of food, specifically a fried egg on top of a bed of rice. The egg is cooked sunny-side up, with the yolk still runny, and it's garnished with sliced red peppers and green herbs. The rice appears to be seasoned with spices, adding a pop of color to the dish. The plate is pink, which contrasts nicely with the vibrant colors of the food. The video is shot in a realistic style, focusing on the textures and colors of the food, making it look very appetizing. The camera angle is slightly above the plate, giving a clear view of the food and its garnishes. The lighting is bright, highlighting the details of the dish. The video does not contain any text or additional elements, keeping the focus solely on the food."
+    ],
+    "video_ids": [
+        "Up6IbNnjzl8_58_340to461"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A fried egg, rice, sliced red peppers, green herbs.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a fried egg with a runny yolk, rice mixed with ground meat and vegetables, sliced red peppers, and green herbs (likely basil). All the specified elements are present and accurately depicted, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a fried egg, a runny yolk, rice, with corn nuts, vegetables, sliced red peppers, and green herbs.likely parsley or The the elements objects are present, match depicted, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4d8c26ee695d97b82743121d14b9d7e2e9d4dede
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f25a371cb0e229c280c58d6b57ee179aa11abf72a80e3547dec5b6805ab54514
+size 83294
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..23c8028a55ca614fb45cfec7e6995e797d1d417e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person holding a partially eaten ice cream cone. The ice cream is white and appears to be vanilla flavored. The cone is brown and has a waffle-like texture. The person is holding the cone with their left hand, and the cone is tilted slightly to the left. The background is blurred, but it appears to be an outdoor setting with a sidewalk. The style of the video is casual and informal, capturing a simple, everyday moment."
+    ],
+    "video_ids": [
+        "LJnuyjpZiwI_15_0to126"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a partially eaten ice cream cone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person holding a partially eaten ice cream cone with a clear plastic straw inserted into it. The cone contains white ice cream with sprinkles and what appears to be a piece of fried food (possibly a potato or similar item) partially eaten, matching the description of 'A person, a partially eaten ice cream cone.'\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close holding a partially eaten ice cream cone. a w view wrapper visible into it. The cone is a ice cream, someles, appears appears to be a green of green food,possibly a fish chip a item) on visible. which the description of aa person, a partially eaten ice cream cone.'\"\n}<|im_end|>\nGuidId",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5b7547f574e06dbbb6672d61969479f42761c914
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e1463320ae7b06e3e4e2a247fb881e80762c3db2bb6f07c4a3f4dc8d0b235f3
+size 87373
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..0eb93cc5fe3f915620af2a28e39360fbec24133f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young woman is seen walking down a tree-lined path. She is dressed in a gray tank top and her hair is styled in a short bob. The path she is walking on is lined with lush green trees and bushes, creating a serene and peaceful atmosphere. The woman appears to be in deep thought as she walks, her gaze directed off to the side. The overall style of the video is natural and organic, capturing the beauty of the outdoors and the quiet introspection of the woman."
+    ],
+    "video_ids": [
+        "O4if2mLgwBs_17_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman, a gray tank top, a short bob hairstyle? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman wearing a gray tank top and has a short bob hairstyle, matching the description. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young woman wearing a gray tank top and a a short bob hairstyle. which the description provided The background is lighting elements in not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f14ad09d92c032ab2e6b75ec212cb98600187d8e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1163-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7ea1889c77b37e960860501edbac11f627ab03dbeac44fcffa13a71714e87e20
+size 143905
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..83f9a2783337b47d262b0521df2e4c2393409e1f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a cozy scene of a campfire in a forest. The fire is burning brightly, with flames licking at the logs and a pot sitting on a grate above the fire. The pot is black and appears to be cooking something. The fire is surrounded by logs, creating a warm and inviting atmosphere. In the background, a yellow sign with black text reads \"BRIDGES FREEZE BEFORE\". The sign is slightly blurred, suggesting it's in the distance. The overall style of the video is natural and rustic, capturing the essence of a camping trip in the woods."
+    ],
+    "video_ids": [
+        "9QqmWuQmk6Q_30_0to154"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Campfire, flames, logs, black pot, yellow sign? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a campfire with visible flames, surrounding logs, a black pot placed over the fire, and a yellow sign in the background that reads 'BRIDGES FREEZE BEFORE'. All specified objects are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a campfire with flames flames, logs logs, a black pot placed on the fire, and a yellow sign in the background. reads 'BUEG RE!!'. R The these objects are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..082ac6f6c0a7cfb8920cb74d405b5c331d99d185
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:14ef82ecaf4a939b8c858bfca69d7ff8d3103e70e349174e91e24ddc74c72245
+size 199734
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..91e7aad80a9ff169377d77ed0aee359e96ab8d0b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a gray hatchback car parked on a tiled plaza. The car is positioned in the center of the frame, with its front facing the camera. The car has a sleek design, with black rims and a shiny exterior. The plaza is surrounded by tall buildings, suggesting an urban setting. The sky is clear and blue, indicating that the video was taken during the day. The overall style of the video is modern and sleek, with a focus on the car and its surroundings."
+    ],
+    "video_ids": [
+        "2-bDhAC1xqA_7_0to123"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A gray hatchback car with a sleek design, black rims, and a shiny exterior.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a gray hatchback car with a sleek design, black rims, and a shiny exterior, matching the description. The car is prominently featured, and while there are additional background elements (buildings, sun, pavement), they do not contradict the core description of the car.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a gray hatchback car with a sleek design, which rims, and a shiny exterior. which the description provided The car's positioned displayed in and the the are additional elements elements likebuildings and sky, and), they do not conflict the core description of the car.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ddd20892de4f3ba3cfbb63dda1bcc68be2c2fb18
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8178772601c5970f35e75206500c95929a5aa2156b75c1eb58275ba544b43638
+size 263671
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..302e3d2a2bc6adb668abf302dbae9def2188d1c9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car, focusing on the dashboard and steering wheel. The dashboard features a digital display with various icons and information, including the time, speed, and fuel level. The steering wheel is black and has multiple buttons on it. The car's interior is modern and sleek, with a focus on technology and convenience. The style of the video is a close-up shot, providing a detailed view of the car's interior. The video is likely meant to showcase the car's features and design."
+    ],
+    "video_ids": [
+        "-IoFtIjWSkA_23_0to198"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dashboard, digital display, icons, information (time, speed, fuel level), steering wheel, buttons? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows the dashboard with a digital display, including icons and information such as time, speed, and fuel level. The steering wheel and buttons are also clearly visible, fulfilling all the specified conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a dashboard, a digital display, icons icons and information such as time, speed, and fuel level. The steering wheel and buttons are also visible visible, fulfilling the the conditions conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..91e2ec2b1b3fae77b07a69f26935f0c61bb2c5e4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5dae83af384c2ae6504c021b8403bceb157c63324a69ef796e6c43e99c95cb1f
+size 135225
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..c7aaf61988d4cea48604a74f2e8a982bcf738506
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a large, brown and beige motorhome parked on a concrete surface. The motorhome has a distinctive design with a curved front and a large windshield. It is equipped with a side mirror and a side door. The motorhome is parked next to a building, and there is another motorhome visible in the background. The sky is clear and blue, suggesting a sunny day. The style of the video is straightforward and documentary, capturing the motorhome in its environment without any additional embellishments or effects."
+    ],
+    "video_ids": [
+        "L8CW33KFE9Q_46_0to113"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large, brown and beige motorhome with a curved front and large windshield, equipped with a side mirror and a side door. Another motorhome is visible in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large motorhome that is brown and beige with a curved front and large windshield, equipped with a side mirror and side door, matching the description. Another motorhome is visible in the background, also matching the description. There are no conflicting elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large,home with is brown and beige in a curved front and a windshield. equipped with a side mirror and a door. which the description. Another motorhome is visible in the background, which align the description. The are no elements elements in contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d19d4d8c7296c07f893654f5f39938baec466d9f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6ec344f652e9c07d3c4ba9733ffa197aee8afde9f0d6a647ccb39f756554d97c
+size 161504
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..68070d352ef9f42c9c3b39f7840461507c95f9c7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene sunrise over a vast, open field. The sky is filled with a dynamic array of clouds, ranging from light to dark gray, with patches of blue peeking through. The sun, positioned low on the horizon, gradually rises, casting a warm, golden glow across the landscape. As the sun ascends, it illuminates the clouds, creating a beautiful interplay of light and shadow. The field below is expansive and mostly barren, with some scattered trees and distant hills adding depth to the scene. A lone horse stands in the distance, adding a sense of tranquility and life to the otherwise still environment. The camera remains stationary throughout the sequence, allowing viewers to fully absorb the peaceful and picturesque moment as the sun continues its ascent, slowly brightening the entire scene."
+    ],
+    "video_ids": [
+        "2ddfa3a22a84715b446d793127545b045d1c222ec1c90b3c7d23c891a94f4147"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Scattered trees, distant hills, a lone horse? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts scattered trees along the horizon, distant hills in the background, and a lone horse visible on the right side of the field. These elements are consistent with the described 'Object(s)' condition, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a trees in the horizon, distant hills in the background, and a lone horse grazing in the right side of the frame. The elements align consistent with the description 'Object(s)' condition.\"\n and there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..69c11587f1bc88595552268daafd04c19fd985ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:406891b9f380d947b91381196177462f9376fd9b66a5ce4366feb4827cdea94b
+size 132937
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..ab06106afcf0b1597ab1b2e266fdee4089848ab8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a blue car on display at a car show. The car is parked on a gray carpet, and its shiny exterior reflects the bright lights of the showroom. The car's black rims add a touch of elegance to its overall design. The car is positioned in the center of the frame, drawing the viewer's attention immediately. The showroom itself is spacious and well-lit, with other cars and booths visible in the background. The video captures the essence of a car show, highlighting the beauty and design of the blue car on display."
+    ],
+    "video_ids": [
+        "WXi3-JyFKz0_41_0to117"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue car, black rims? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a blue car with black rims, which matches the core description. The car is prominently featured and the details of its color and wheel design are visible throughout the frames.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a blue car with black rims. which matches the description description. The car is prominently displayed in the color of its color and rims design are visible, the frames.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fa3c2b61be6bbe21e758caec29b5d254fac3d8c4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4bffbbc7144e58d85b4448a0df43d5d2810cb58925da13c54e56a9c9ef90a2e1
+size 228179
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..9027fafc6fbb40fac40a3f8b8048797283b1ebc4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a playful and colorful scene featuring a variety of toys and figurines. The scene is set on a vibrant, multicolored play mat with a whimsical design. In the foreground, there is a large, cartoonish figurine of a man with a red shirt and brown pants, standing with his arms crossed. Behind him, there are several smaller figurines, including a group of colorful ponies with different mane colors and styles. To the right, there is a red toy house with a clock on its side, and a toy bed with a pillow and blanket. The overall style of the video is playful and child-friendly, with a focus on bright colors and fun, cartoonish designs. The toys are arranged in a way that suggests a playful, imaginative scene, possibly representing a child's playtime."
+    ],
+    "video_ids": [
+        "Z_2UIRHOOKU_19_0to135"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large, cartoonish figurine of a man, several smaller figurines (including colorful ponies), a red toy house, a toy bed with a pillow and blanket.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large, cartoonish figurine of a man (Wreck-It Ralph), several smaller figurines including colorful ponies (My Little Pony), a red toy house (with a clock and bookshelf), and a toy bed with a pillow and blanket (with a character lying in it). All elements described in the condition are present in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a large, cartoonish figurine of a man standingtheario-It Ralph) several smaller figurines ( colorful ponies,from Little Pony), a red toy house,a a clock on a on), and a toy bed with a pillow and blanket.in a colorful lying on it). The these match in the condition are present in the video.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9307c4ed08a44de06cbf01ac8ab03636029f85ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6884a0ae75832ce445d2ac9f5d42324069a215a2684b842713682fd6d9a25b41
+size 84227
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..ca6000985b5fdb0eb73c6613c016bf5d17461766
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a gray and white cat sitting on a concrete surface. The cat has a fluffy coat and is looking directly at the camera with a focused gaze. The cat's ears are perked up, indicating alertness. The cat's tail is curled around its body, and its whiskers are spread out. The background is blurred, but it appears to be an outdoor setting with some debris scattered around. The style of the video is a close-up shot with a shallow depth of field, focusing on the cat's face and upper body. The lighting is natural, suggesting it might be daytime. The overall mood of the video is calm and serene, with the cat appearing relaxed and comfortable in its environment."
+    ],
+    "video_ids": [
+        "DehYCPtoGcI_11_0to192"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A gray and white cat? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a gray and white cat with distinct tabby markings, consistent with the description. The cat's fur is primarily gray with white accents on its paws and face, and it has the characteristic appearance of a domestic cat. There are no elements in the video that contradict the description of a gray and white cat.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a cat and white cat sitting a markingsby markings, sitting with the description of The cat is fur pattern primarily gray with white patches, its chestaws and chest, and it has the characteristic tab of a tab tab. The are no additional in the video that contradict the description of a gray and white cat.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0dcb1e8dbcb13e651225c7c75d6853e88aa244b1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1164-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48b4017d9ebc92a741fd897287aab56d88d6088a1845041caeb5fb3a8b0c4072
+size 93819
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..b003a1dece7e47cf20382c06c956ff99d857f232
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a bouquet of flowers in a pink vase placed on a white surface. The bouquet consists of various types of flowers, including pink, red, and yellow ones, as well as some greenery. The flowers are arranged in a way that they fill the vase and spill over the edges slightly. The vase is positioned in the center of the image, and the background is a plain white wall. The style of the video is simple and straightforward, focusing on the beauty of the flowers and the simplicity of the setting. The lighting is soft and even, highlighting the colors of the flowers without creating harsh shadows. The overall impression is one of tranquility and natural beauty."
+    ],
+    "video_ids": [
+        "1CegHAsM7bw_5_0to140"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bouquet of flowers in a pink vase.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a bouquet of colorful flowers (pink, red, and yellow) arranged in a pink vase, which matches the core description. Additional elements like a skull and candle are present but do not contradict the main object described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a bouquet of flowers flowers arrangedpink, yellow, and yellow) arranged in a pink vase. which matches the description description of The elements like the white and a are not but do not conflict the main object description.\"\n}<|im_end|>\nGuidId",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..64e8292323a41228a44c58a580c8cf86abbf98ee
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:530b1587af5576d4d0bbf0ac058ff47f6f9495871b6a1f3f60b6a8363601c689
+size 67642
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..cedfc6f43fef29e707b2d61860139e4bc1f1fa58
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen sitting on a couch with white pillows and a white curtain in the background. She is wearing a red hoodie and has her hair pulled back. In her hand, she holds a small blue and white object, which appears to be a toothbrush. She is looking at the object with a thoughtful expression on her face. The overall style of the video is casual and relaxed, with a focus on the woman and her interaction with the object."
+    ],
+    "video_ids": [
+        "GPq9nKkkgpc_31_0to103"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a couch, white pillows, a white curtain, a small blue and white toothbrush, the woman's red hoodie, and her pulled-back hair.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman wearing a red hoodie with her hair pulled back, sitting on a couch with white pillows. Behind her is a white curtain with string lights. She is holding a small blue and white toothbrush, which is visible in her hand. All elements described in the condition are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a woman wearing a red hoodie, her hair pulled back, sitting on a couch with white pillows. She her, a white curtain. a lights, She is holding a small blue and white objectbrush. which align consistent in her hands. The the described in the condition are present and match with the video content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3ec6f1d949dadc9f6b81a05f32691b87cc778fe1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eaccfdcd8a0c3a569d2ca908f3f85edd0c1e71be4dddcd39fa5c38c3e5e6c1c9
+size 172018
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..4d556a5530c97896791615986d37977e16efc241
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a young girl in a shopping mall, her face framed by a blue headband. She is seen in three different frames, each capturing a different moment of her journey. In the first frame, she is seen walking towards the camera, her eyes focused on something in the distance. In the second frame, she has stopped and is looking up, her gaze fixed on a sign or a display. In the third frame, she is seen walking away from the camera, her back turned to the viewer. The mall around her is filled with people and shops, creating a bustling atmosphere. The girl's journey through the mall, from one frame to the next, tells a story of exploration and discovery."
+    ],
+    "video_ids": [
+        "IkKzqU5Cq7E_64_0to155"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young girl? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a young girl as the main subject, who is clearly visible and central to the frame throughout the clip. She is engaged in the action of eating, which aligns with the description. While there are other elements in the background (e.g., people, tables, mall environment), they do not contradict the core description of a young girl being the primary focus.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a young girl wearing the main subject. wearing is consistently visible in occupies in the frames in the sequence. The is wearing in a environment of looking, which iss with the ' of The there are other elements in the background,such.g., a, lights, and environment), they do not detr the core description of the young girl eating the main focus.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ae24bfc942b1a0d3f78eee07556ce68a98c6b466
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf5b569c59a0f88150e6fa31124a1777382ef117232c7323c9f620e9c713a810
+size 192391
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..40366c317ca00f2e8ba421b69e021c45c5fd887f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and a scarf, standing in front of a television screen. The man appears to be speaking or reacting to something on the screen. The television screen displays a colorful image, possibly a video or a slideshow. The man is dressed in a casual manner, with a scarf wrapped around his neck. The setting seems to be indoors, possibly a living room or a media room. The overall style of the video is casual and informal, with a focus on the man's reaction to the content on the screen."
+    ],
+    "video_ids": [
+        "BcchAQjOq-E_17_0to130"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and a scarf, a television screen displaying a colorful image.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man with a beard and a scarf standing in front of a television screen that is displaying a colorful image. The core elements described are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man with a beard and a scarf, in front of a television screen. displays displaying a colorful image. The description elements of in present, accurately depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d9270108c39252944c6fcbc132cd83b760b9c329
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b75add2caf9f2b47bcec6b7b2f7da4f04a7c80d057af8898b173ccbccefd51ee
+size 186107
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..0e08862318fd231a0797d1132203ef56aa806f2f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a chef in a professional kitchen, preparing a dish. The chef is wearing a white chef's coat and is focused on stirring a pot of food on a stove. The stove is a modern, stainless steel model with a digital display and control panel. The chef is using a long-handled spoon to stir the contents of the pot. On the counter next to the stove, there are several bowls and cups, suggesting that the chef is in the process of preparing multiple ingredients for the dish. The kitchen is well-lit and clean, with a professional and organized appearance. The style of the video is a straightforward, documentary-style recording of the chef's actions, with no additional embellishments or artistic effects."
+    ],
+    "video_ids": [
+        "IGxlQfyjK2Y_3_24to182"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chef, stove, long-handled spoon, bowls, and cups.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chef in a white uniform using a stove with a pot, stirring with a whisk (a type of long-handled spoon), and there are bowls and cups on the counter. All the specified objects are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person stirring a kitchen uniform stirring a long to a long on stirring with a long,which long of long-handled spoon), and there are bowls and cups visible the counter. The the elements objects are present and the with the description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..314bd3b43e66181de294635ac192d04a030a628c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b45fa65ac88849e744078f0244e6ba73cbed5fd8f01529fbb848907a648951fa
+size 182207
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..4f69376faee98254092740828ef100983cefd12d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person using a fork to pick up food from a plate. The fork is silver and has a long handle. The food on the plate appears to be a type of meat, possibly chicken or beef, and it is covered in a sauce. The plate is white and round. In the background, there is a bowl that contains more of the same food. The person is holding the fork with their right hand and is using it to pick up a piece of the food. The style of the video is a close-up shot, focusing on the fork and the food on the plate. The video does not show any other objects or people. The focus is solely on the person using the fork and the food on the plate."
+    ],
+    "video_ids": [
+        "b7xSUhljnY0_37_23to195"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: - Fork? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a fork being used to assist in lifting food with a spoon, fulfilling the 'Fork' condition. The fork is visible and actively engaged in the action, consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a fork being used to interact in picking a from a spoon. which the 'Objectork' condition as The fork is visible and actively engaged in the action, which with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..945ae2add4e87de4883c92b22794f42629f4f62f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:741066df7619e63c553e194ab8b72aea9c6aea821e3cc3256999ed8c8375459e
+size 145700
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..02e72c1f197e222cce91ae620cb8f6f693a347ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a conversation between two women in a room with a bookshelf in the background. The woman on the left is wearing a pink top and has her hair pulled back. She is speaking and appears to be engaged in the conversation. The woman on the right is wearing a white top and has her hair styled in a bob. She is listening attentively to the other woman. The room has a warm and inviting atmosphere, with the bookshelf adding a touch of intellectual charm. The women seem to be discussing something important, as indicated by their focused expressions."
+    ],
+    "video_ids": [
+        "X6DRLXo9tq8_46_0to172"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two women? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two women engaged in a conversation. One woman is in focus, wearing a pink top, and the other is partially visible, wearing a pink cardigan. Their presence and interaction align with the 'Object(s): Two women' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two women sitting in a conversation. They is is wearing a, wearing a pink top, while the other is in visible, wearing a white topigan. The presence and interaction fulfill with the descriptionTwo(s)' Two women' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5f200f4d03d7734adbb4547bdbb97b0cc127a2e5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d047d461624ce4c1d96458e1dbe523dfcd9d1c604e475f78ff342e69ed1ee8e0
+size 95783
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..a45460b43baff894a8a10bc786b30814f139c576
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a journey of a boat on a river. The boat, painted in a vibrant shade of blue, is equipped with a yellow roof, providing shelter from the elements. The boat is adorned with a large speaker system, hinting at the possibility of music or announcements during the journey. The boat is seen moving along the river, passing by lush green trees that line the riverbank. The journey appears to be peaceful and serene, with the boat cutting through the calm waters of the river. The overall style of the video suggests a leisurely river cruise, with the boat serving as the main focus, surrounded by the natural beauty of the river and its surroundings."
+    ],
+    "video_ids": [
+        "1BPQt65_Muw_32_0to129"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A boat painted in vibrant blue with a yellow roof and a large speaker system.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The boat in the video is painted in a vibrant teal/blue color with a yellow roof, matching the description. It also features a large speaker system with multiple horns visible on the side. The presence of other elements like the cabin, lights, and surrounding environment does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video in the video is indeed in vibrant vibrant blue blue color with a yellow roof, which the description. Additionally also features a large speaker system, multiple speakers, on the top, The overall of the elements like the water windows windows, and the environment does not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..352369a06500271827e9756aab3d866f3897aa22
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1165-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:91edba135578c3db4f4a67979d9df0d33eaa62302ca759be6b0df2d4471ef542
+size 144913
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..2039d14c11ce1e4b0833358d236b5ba6ae1a75b4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a military uniform standing in a room filled with people. The man is bald and has a mustache. He is wearing a black jacket with medals on it. The room has white balloons and blue balloons. There are also chairs and tables in the room. The man is looking at the camera. The style of the video is a real-life event captured in a candid manner."
+    ],
+    "video_ids": [
+        "4A38ZhjL1rU_12_36to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a military uniform, black jacket with medals, bald with a mustache. The man is looking at the camera.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a military uniform with a black jacket adorned with medals, and he has a bald head with a mustache. He is looking directly at the camera throughout the frames, matching the core description provided. Additional elements like balloons and background attendees do not contradict the main subject's appearance or action.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a black uniform, a black jacket adorned with medals, and he is a bald head with a mustache. He is looking directly at the camera, the video, which the description description provided.\"\n The elements like the and a decorations do not contradict the main description description description and actions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6bbd735ca2152eb2652b0f425f894a5c8abca544
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6501cdcc21afdb5ad36f824d10807804be804200f76759cd50449197ec235972
+size 149219
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..7b8fd7d9a8d30ba9d19ce0830768abd893e90c17
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dramatic scene where a small, colorful toy resembling a bird is placed under a hydraulic press. The toy, primarily yellow with blue accents on its head and wings, sits on a metallic platform. As the press descends, it begins to compress the toy, causing it to deform and eventually burst apart. The toy's body splits into two halves, revealing its internal structure, while the blue feathers and yellow body parts scatter around the platform. The hydraulic press, marked with black and yellow stripes for safety, exerts immense pressure, flattening the toy completely. The background features a green wall with some visible wear and tear, adding an industrial feel to the setting. The camera remains stationary throughout, focusing closely on the interaction between the press and the toy, highlighting the transformation from intact to fragmented."
+    ],
+    "video_ids": [
+        "aac6d9e420e4b316685531b05ac4d1cf5f5d6095309adfe2f8bb90c8db445c2a"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small, colorful toy resembling a bird, a hydraulic press, and debris from the toy\u2019s fragments.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small, colorful toy resembling a bird (blue and yellow with a cartoonish face) being crushed by a hydraulic press. As the press descends, the toy deforms and breaks apart, creating visible debris that spreads around the press base. All core elements described are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a small, colorful toy resembling a bird beingyellow and yellow) a yellowish design) being crushed by a hydraulic press. The the press descends, the toy isforms and eventually into, creating debris debris from sc around the base.. The elements elements of in present: accurately depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a79169e160b566fe04e52193de05630ad0871f6d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6fae8c2f5302da745af3ab8379b3aaf44dc84df160412cffa296b8ac745774d2
+size 207445
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..294c5e5bd1562ee268cf5b76f90d34880d999a90
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a bustling city street at dusk, with the sky transitioning from day to night. The street is wet from recent rain, reflecting the city lights and adding a glossy sheen to the scene. Pedestrians are seen walking along the sidewalk, some carrying umbrellas, while others are huddled under awnings. The traffic is light, with cars moving in both directions. The architecture of the buildings is varied, with modern high-rises and older, more traditional structures. The overall atmosphere is one of a vibrant, urban environment, with the city's lights beginning to take over as the day gives way to night."
+    ],
+    "video_ids": [
+        "cD56SFItbGQ_270_0to104"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Pedestrians, cars, awnings? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows pedestrians (a person with a red umbrella crossing the street), cars (multiple vehicles moving through the intersection), and awnings (visible on buildings along the street). All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful pedestrians walkingpeople couple holding an blue umbrella and the street), cars (a cars on along the street), and awnings (visible on the to the street). The these elements of in present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..73e0abb9abec05bc2d63a22d56ab6fb0e00db9cb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77db27f120374a7f8828fd9576c48bbe58218b94f8df4554a1271fa0c89dd03f
+size 174637
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..7b0d48554b3393c72872d4bf0cbcc148ef211750
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features two animated dolls, one with blonde hair and the other with red hair, standing in front of a closet. The dolls are dressed in colorful outfits, with the blonde doll wearing a blue dress and the red-haired doll in a blue and black dress. They are positioned in front of a white door with a window, which is slightly ajar. Inside the closet, there are various items, including a pink jacket and a green bag. The dolls appear to be engaged in a conversation, with the blonde doll gesturing towards the closet. The overall style of the video is playful and colorful, with a focus on the dolls and their interaction with their surroundings."
+    ],
+    "video_ids": [
+        "SZguwQ9hHQQ_8_746to918"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two animated dolls, one with blonde hair and the other with red hair, dressed in colorful outfits. The blonde doll wears a blue dress, and the red-haired doll wears a blue and black dress.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two dolls that match the description: one with blonde hair in a blue dress (Elsa) and one with red hair in a blue and black dress (Anna). Their outfits and hair colors are accurately represented, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two animated with match the description provided one with blonde hair and a blue dress andwithlsa-like and the with red hair in a blue and black dress (Anna). The outfits and hair colors are consistent represented, and the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c7b8bbe3286778b0152f6d5270a6985cb4054e67
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:08201c4da8284d275750248469d2e3fe98a890f90c5bf2849516b1d3ecbefcf8
+size 278473
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..eaff65d8bb724db5f6315c6ef722cafe89469ea6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment on a soccer field. A soccer player, dressed in a light blue jersey and white shorts, is in the midst of a triumphant celebration. He is pointing towards the sky, his arms raised high in victory. His joy is palpable, reflecting the excitement of the game. The field is a vibrant green, contrasting with the player's light blue jersey. In the background, other players can be seen, their attention focused on the celebrating player. The image is filled with the energy and passion inherent in the sport of soccer."
+    ],
+    "video_ids": [
+        "pyKescKQnbU_64_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A soccer player in a light blue jersey and white shorts, and other players in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player in a light blue jersey and white shorts, celebrating with arms outstretched and pointing. Other players in similar uniforms are visible in the background, consistent with the description. The scene is set on a soccer field with spectators, which aligns with the context of a soccer match.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player in a light blue jersey and white shorts, which with his raisedstretched. a upwards In players in similar uniforms are visible in the background, standing with the description. The setting is set on a soccer field, a, which aligns with the context of a soccer match.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..afdcbfcd985cc27fe4fbf530847a3d0a68480244
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f2510b1bce282ab261d31b82615bd2a8895860922699053d198ca7d853833135
+size 148331
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..3897a6b1c867fd7b99785fd9032b08e23e2a3bec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a man in a blue suit sitting in front of a window with a cityscape view. He is speaking and appears to be discussing a sports-related topic. The screen also displays a graphic with a football player and statistics, indicating that the topic is related to the NFL. The style of the video is a typical news broadcast with a focus on sports reporting. The man is the main subject, and the background provides context to the location of the broadcast. The overall tone of the video is professional and informative."
+    ],
+    "video_ids": [
+        "GKMLibhRs-U_2_644to790"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a blue suit, window, cityscape, graphic with a football player and statistics? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a blue suit speaking, with a window and cityscape visible in the background. A graphic at the bottom includes a football player image and statistics, matching the specified 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man in a blue suit, in with a window in cityscape visible in the background. There graphic with the bottom of a football player and and some, which the description elementsgraphic(s)'.\"\n.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..662b16da552d72f2216f4a8f29dfedea4aa69e35
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0db88e5ca8c3f94ef4fd1cd084ab8322c749f8e4ce861efe6e95f19b8e31deb5
+size 131047
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..c7ec6eb38f3395a425cd33a64ac2bc228bcbd836
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with blonde hair, wearing a gold necklace and a beaded dress, sitting at a table with a menu in front of her. She is holding a glass of water and appears to be speaking. The setting is a restaurant with a blue and white color scheme, and there are wine glasses and a bottle on the table. The style of the video is a candid, real-life scenario, possibly from a television show or a documentary."
+    ],
+    "video_ids": [
+        "FU1NKB3Ge20_0_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with blonde hair, wearing a gold necklace and a beaded dress, holding a glass of water and sitting at a table with a menu in front of her. She appears to be speaking.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair, wearing a gold necklace and a beaded dress, sitting at a table with a menu in front of her. She appears to be speaking, and there is a glass of water on the table. Additional elements like a trophy and a person holding a box are present but do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a woman with blonde hair wearing wearing a gold necklace and a beaded dress, holding at a table with a menu in front of her. She is to be holding and holding she is a glass of water in the table. The elements like the yellow and a yellow in a drink in present but do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c32ccafaf8533f3a313b2018529ff6cd0de0f03e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3833ca626344ee2543760270531d8d33358322b6136538c9cb1f9d5c4527759
+size 138916
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..e956af1feccf8031d97427d008b8e59bd27f9a8c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a plate of food, specifically a dish of tacos, being served on a green table. The tacos are topped with various ingredients, including shredded meat, chopped onions, cilantro, and a slice of lime. The food is presented on a white plate, which contrasts with the green table. The style of the video is a simple, straightforward food presentation, with no additional elements or actions. The focus is solely on the dish of tacos, highlighting the colors and textures of the ingredients. The lighting is bright, making the food look appetizing and fresh. The video does not contain any text or additional elements, and the camera angle is a close-up shot of the plate of food."
+    ],
+    "video_ids": [
+        "0nDqNnLZofs_23_0to154"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Tacos, shredded meat, chopped onions, cilantro, slice of lime, white plate? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a white plate with tacos topped with shredded meat, chopped onions, cilantro, and a slice of lime, matching the described elements. Additional elements like the green table and chairs do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a plate plate with a, with shredded meat, chopped onions, cilantro, and a slice of lime. which the description elements. The elements like the green backgroundcloth the in not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..847442b0c18ab60abe5d1d3721cf34579a21bc5d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1166-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:107c717ad2749190b3bd2e2d200db29bff422f4bd7aed889d85d323ef3392cc5
+size 104691
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..0b096011662e38153dba96e0f84a60e129f36bee
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a cozy scene of two cups of coffee and four cookies on a white plate. The plate is placed on a blue napkin, which contrasts with the warm tones of the wooden table. The cookies are round and have a golden-brown color, suggesting they are freshly baked. The coffee cups are white with a handle on one side, and they are filled with a dark liquid, likely coffee. The arrangement of the items on the plate and the cups suggests a casual and inviting atmosphere, perhaps for a morning or afternoon break. The overall style of the video is simple and clean, focusing on the objects and their arrangement rather than any action or movement."
+    ],
+    "video_ids": [
+        "zv6g6RVMV_M_8_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two cups of coffee and four cookies.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows two cups of coffee and four cookies. Initially, cookies are placed on a cooling rack, and later they are arranged on a plate alongside two cups of coffee, matching the described objects without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful two cups of coffee and four cookies. The, the are placed on a plate rack, and then they are transferred on a plate. the cups of coffee. which the description objects. any.\"\n}<|im_end|>\n\ufffd\ufffd",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f43e015ee3465b32f15408a683a10ce6fe90dd52
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c7a46c3bb685a508cc089ceb4b2e9216c70b7c1bf25f5310cd53a77506e070a7
+size 78624
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..1619c9f4e2c8b8a851acb26d1376c7a09dad7a71
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures an intense moment in a football game. The main focus is on two players, one in a red jersey and the other in a white jersey. The player in the red jersey is attempting to block the player in the white jersey, who is in the process of catching the football. The action takes place on a football field, with other players visible in the background. The style of the video is dynamic and fast-paced, capturing the high-energy atmosphere of the game. The camera angle is from the side, providing a clear view of the players' actions. The video does not contain any text or additional elements."
+    ],
+    "video_ids": [
+        "_WdZFPYIDis_7_0to159"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two players, one in a red jersey and another in a white jersey.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two football players: one wearing a red jersey (number 20) and another wearing a white jersey (number 11). These two players are the central focus of the frame, fulfilling the 'Object(s)' condition as described. Additional elements like spectators and officials in the background do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts two players players, one in a red jersey andnumber 32) and the in a white jersey (number 66). The players players are engaged main focus of the video, fulfilling the 'Object(s)' condition described described. The players, the and other in the background do not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0b956eb19645d4f3df0a93575011221a2ca1ba93
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3532b5f31311a7c2b4dd103b2045523a6195ff94e6bcc31bce845f277af3a94f
+size 343065
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..313bdf7fa83434579500812f59faf1bc36edeae1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is sitting on the floor, surrounded by a large pile of blue and gray toy train tracks. The person appears to be sorting through the tracks, possibly organizing or assembling them. The tracks are scattered around the person, with some pieces overlapping and others lying flat. The person is holding a small blue piece of track in their hand, which they seem to be examining or picking up from the pile. The setting appears to be indoors, with a wooden floor beneath the person and a white wall in the background. The overall style of the video is casual and candid, capturing a moment of play or work with the toy train tracks."
+    ],
+    "video_ids": [
+        "YXFhzhbNaSc_5_24to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, blue and gray toy train tracks, a small blue piece of track.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person sitting on the floor surrounded by blue and gray toy train tracks, and the person is holding a small blue piece of track. These elements match the description exactly, and additional elements (like other toys or packaging) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a person interacting on the floor surrounded by blue and gray toy train tracks. which the person is interacting a small blue piece of track. The elements match the description of, fulfilling there elements likelike the colors or the) do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d87dc067129a4333e283e81281533ee2ab953a1e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40a974de8b9fe04834ecf0bbfcc64aee5e6d470b5d8ae2db91a7e70ab440dc3c
+size 220799
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..06929e4ef395e01d9d23904538ec170e9253d241
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a sea turtle swimming in the ocean. The turtle, with its brown and white shell, is seen in three different positions as it moves through the water. The turtle's head is visible in the first frame, while its tail is seen in the second frame. The third frame shows the turtle swimming away from the camera. The ocean floor, visible in the background, provides a contrasting backdrop to the turtle's movement. The video is a beautiful representation of marine life, showcasing the grace and beauty of the sea turtle in its natural habitat."
+    ],
+    "video_ids": [
+        "_vRyBq7poEg_23_0to205"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Sea turtle? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a sea turtle swimming underwater, which aligns with the 'Object(s)' condition of 'Sea turtle'. The presence of a remora fish on its back does not contradict this, as it is a common natural occurrence and does not negate the primary subject being a sea turtle.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features sea sea turtle swimming underwater. which directlys with the 'Object(s)' condition. theSea turtle'. The presence of another secondora fish in the shell is not contradict the description as it is a common behavior occurrence for does not change the primary subject being a sea turtle.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0b851a6704eff61186989cff837f7beab167b55c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a7cede9161039ca19bdd69a604d1b0b4491eb881491d08dcbec49d640c91a92e
+size 325501
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..f80e9fa96ae0607052b48d030871f9f262de078a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a playful and imaginative scene set in a suburban neighborhood. It features two green dinosaurs, each with a pair of blue sunglasses, walking down a grassy lawn. The dinosaurs are anthropomorphized, with one of them holding a red guitar, suggesting a sense of fun and adventure. The background shows a typical residential area with houses and a fence, providing a contrast to the whimsical dinosaurs. The video is likely aimed at a younger audience, given the cartoonish style and the playful elements."
+    ],
+    "video_ids": [
+        "QVcsnQW0CT0_61_119to327"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two green dinosaurs wearing blue sunglasses, one holding a red guitar.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two green dinosaur-like creatures wearing blue sunglasses. One of them is holding a red guitar, matching the specified condition. The background elements (houses, grass, etc.) do not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two green dinosaurs-like characters wearing blue sunglasses. One of them is holding a red guitar, which the description objects. The setting and,houses, fence, and.) do not contradict the description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..454da08d4bcf4cc0aeceefbf300edf9c9fd1f219
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75c5a35a95022ab1843a22f645ec78fac9aefee22b486c9b9814222691a3a136
+size 285248
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..0dc5918c5b94fcf5185285ed575deb4ef6db4900
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man with a mustache and a black afro hairstyle is seen sitting on a couch. He is wearing a black leather jacket and a colorful jacket with orange and green circles. The man is looking directly at the camera with a serious expression. The background is a blue room with a white wall and a circular window. The man's position on the couch and his direct gaze at the camera suggest that he is the main subject of the video. The overall style of the video is straightforward and focused on the man."
+    ],
+    "video_ids": [
+        "_t5vzglGER4_29_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a mustache and black afro hairstyle sitting on a couch, wearing a black leather jacket and a colorful jacket with orange and green circles.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a mustache and a black afro hairstyle sitting on a couch. He is wearing a black leather jacket with a colorful design featuring orange and green circles on the front. The description matches the visual content accurately, with no significant contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man with a mustache and a black afro hairstyle, on a couch. He is wearing a black leather jacket over a colorful jacket underneath orange and green circles underneath the sleeves. The description matches the visual elements of, with no contradictions contradictions or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b95ed7f79ad81e66b70906722fec45875cb1d292
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b974771a7655ad12d1614202d78979f921cb3ad767ecf011d78810c8ac2feb3
+size 91703
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..1a3aaee766fb18aff83f524f8b71dcf9f7056dac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated scene featuring three anthropomorphic animals in a room with a rustic, industrial feel. The first animal is a white rabbit with pink eyes and a pink headband, wearing a pink jacket with a red collar. The second animal is a brown mouse with a green knitted hat and a brown jacket with a red collar. The third animal is a large brown rat wearing a blue jacket with a red collar and a pair of glasses. The room has a yellowish-brown color scheme with a wooden floor and walls, and there are various mechanical and musical instruments in the background. The style of the animation is detailed and realistic, with a focus on the textures and colors of the animals' fur and clothing. The scene appears to be set in a workshop or a music studio, and the animals seem to be engaged in a creative or musical activity."
+    ],
+    "video_ids": [
+        "7741SqUwajI_0_277to407"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three anthropomorphic animals: a white rabbit, a brown mouse, and a large brown rat.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three anthropomorphic animals: a white rabbit wearing a pink outfit and headband, a brown mouse wearing a green knitted hat, and a large brown rat (or mouse) wearing glasses and a vest. All three match the description, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features three anthropomorphic animals: a white rabbit, a pink outfit, aband, a brown mouse in a green hatitted hat and and a large brown rat wearingor possibly) wearing a and a blue. The three animals the description of and there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0441e0f933c20412544b3b15f7562ee804a70c51
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:679b30bc92a5219def99823c8639b6b5ddf9591c1f7ad170e2bdbc569fb262a5
+size 189135
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..1e1567b76d2512d8ed39b15568d349e2ef36cb4a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a first-person perspective from inside a car, showing the driver's hands on the steering wheel and the view through the windshield. The driver is wearing glasses and appears to be focused on the road ahead. The car is on a city street with other vehicles and buildings visible in the background. The style of the video is realistic and it captures the everyday experience of driving in an urban environment."
+    ],
+    "video_ids": [
+        "G9Rh2wMf5mc_59_87to262"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Driver's hands on the steering wheel, driver wearing glasses, road ahead? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the driver's hands on the steering wheel, the driver wearing glasses, and the road ahead visible through the windshield. All elements of the 'Object(s)' condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a driver's hands on the steering wheel, the driver wearing glasses, and the road ahead. through the windshield. The elements match the 'Object(s)' condition are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8a0e9d611dadb8bf52b90af1ffbb335053268868
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1167-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dfe31c08f6e179016b7d28fdca7a2263a30ae81d7a86bbc8875d747655676959
+size 238800
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..3d871d9d83724d085e372c203c5390da0a6557b9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a scene of a silver Porsche sports car parked on a brick road. The car is sleek and shiny, reflecting the sunlight. The car is parked next to a red sports car, creating a striking contrast between the two vehicles. The brick road adds a touch of rustic charm to the scene. In the background, there are several trucks and a tent, suggesting that this might be a car show or a similar event. The overall style of the video is dynamic and vibrant, capturing the essence of the car show and the excitement of the attendees."
+    ],
+    "video_ids": [
+        "22vjt9ZJv3Q_2_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Silver Porsche, red sports car, trucks, tent? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a silver Porsche in the foreground, a red sports car parked behind it, several trucks in the background, and a white tent on the left side. All elements mentioned in the condition are present and clearly visible, fulfilling the 'Object(s)' requirement.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a silver Porsche and the foreground and a red sports car in behind it, and trucks in the background, and a tent tent. the left side of The these match in the condition are present in do visible in fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f4e99118704c6cc1523d235c7508ea8986025fa1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7fbf569ec5d817f3848562e8dbef63e55ba750b0b1393f2c21ecdf332a287a46
+size 205992
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..c14f29857f642d36b59000300b72d38399de8e73
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a cooking show featuring three hosts standing behind a white counter. The counter is adorned with various food items and cooking utensils. The hosts are engaged in a lively conversation, with one of them holding a book. The setting is a studio with a blue background, and the overall style of the video is professional and informative. The hosts are dressed in casual attire, and the lighting is bright, highlighting the food and the hosts. The video is likely aimed at viewers interested in cooking and food."
+    ],
+    "video_ids": [
+        "s2I0i5pMLq0_1_40to261"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three hosts, various food items, cooking utensils, and a book.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three hosts (a chef, a woman, and a man in a suit) standing behind a counter with various food items (burgers, skewers, etc.), cooking utensils (a pan, bottles), and a book (a tablet or book-like object held by the man). All elements described in the condition are present, and no conflicting elements are visible.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three individuals standingtwo man and a woman, and a man) a suit) standing behind a counter with various food items suchvegeters, vegetablesers, vegetables.) cooking utensils (a knife), a, and a book.possibly tablet). a-like object). by the man in The elements match in the ' are present, fulfilling the additional elements are introduced.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b5b202205e95132c7a6ac071cfba082a58f591e7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0104959235257586fb61fd4ecb3e2369ae9a0506357b3fe6aedf501e8af048c6
+size 136880
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..fbfbd29a1b8a7f9bfe80fe32089ea57cfeef9530
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features Amy Beatie, the Executive Director of the Colorado Water Trust. She is standing in a lush, green forest with trees and foliage surrounding her. The sunlight filters through the leaves, creating a warm and inviting atmosphere. Amy is wearing a green shirt, which complements the natural environment. Her sunglasses are perched on top of her head, and she is looking directly at the camera with a slight smile. The video is likely an interview or a promotional piece for the Colorado Water Trust, highlighting Amy's role and the organization's mission."
+    ],
+    "video_ids": [
+        "3EshDJOVSk4_8_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Amy Beatie (Executive Director of the Colorado Water Trust), green shirt, sunglasses? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows Amy Beatie, identified as Executive Director of the Colorado Water Trust, wearing a green shirt and with sunglasses resting on her head. All core elements of the description are accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a Beatie, who as the Director of the Colorado Water Trust, wearing a green shirt and sunglasses sunglasses on on her head. The the elements of the description are present represented in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b3bdcd2555c4aa97c461ab27983314f1e83234ea
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6bb72b1168947651e97bc195e1c608b462e4def53ad4093e067f47ed150103b
+size 178430
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..569696eff8f2fbaa3e5e3739f27c5d3148e8b598
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman with long dark hair is seen in a conversation with another woman. She is wearing a white blouse with lace details and has her hand on her chin, indicating a thoughtful or contemplative expression. The setting appears to be an indoor space with a purple background, possibly a room with a window or a wall painted in that color. The other woman, whose back is to the camera, is wearing a blue top and is facing the woman with dark hair. The overall style of the video suggests a casual, intimate conversation between the two women, possibly in a personal or professional setting."
+    ],
+    "video_ids": [
+        "8Zhjsj8kTnE_6_0to158"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two women, one with long dark hair wearing a white blouse with lace details, the other facing her with a back to the camera wearing a blue top.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two women as described: one with long dark hair wearing a white blouse with lace details, and the other facing her with her back to the camera wearing a blue top. The core description is accurately fulfilled, with no conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two women. described. one with long dark hair wearing a white blouse with lace details, and the other facing her with a back to the camera wearing a blue top. The setting elements is accurately represented, and no additional elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2fd94ff47cfe27b00a97f51f749d510a1f5ea0b0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fdb60bd148c033b4c8a49f01e9953c3208f09db6aa845249aba9ab7961f29019
+size 177116
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..f39194e89391973132a995aa637c7d1dcf04c994
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a black and white dog drinking water from a clear plastic container. The dog's head is partially submerged in the water, and its tongue is extended, visibly licking the surface to drink. The container appears to be outdoors, as grass can be seen through the bottom of the container. The dog's fur is wet, indicating it has been drinking for some time. The camera remains stationary throughout the sequence, focusing on the dog's face and the water inside the container. There are no significant changes or movements in the scene; the primary action is the dog drinking water."
+    ],
+    "video_ids": [
+        "c8fa649f1ac237d77c80ba8779dd53d7f6f26e0ed723a32863df903b4140a4de"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black and white dog, a clear plastic container? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black and white dog drinking water from a clear plastic container, which matches the described objects. The focus is on the dog's head and the container, with no conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black and white dog interacting water from a clear plastic container. which matches the description '. The presence is on the dog and interaction and the container, and no additional elements present}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5cf7a31b5a90ae3e481323818801eaa3039262e2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd6857e6802be54fe0a41d6e64a40830f863098617068b909997833de3732337
+size 238151
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..0e1093c81bd57e57a543db8ca79382e30ff78e87
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young man with short brown hair is captured in three frames. He is wearing a white t-shirt and a dark gray hoodie. In the first frame, he is looking directly at the camera with a slight smile on his face. In the second frame, he is looking to the left with a thoughtful expression. In the third frame, he is looking back at the camera with a serious expression. The background is a plain gray wall, providing a neutral backdrop that allows the viewer to focus on the man's expressions and attire. The video has a simple and straightforward style, with no additional objects or actions to describe."
+    ],
+    "video_ids": [
+        "T9LHUFwRIK0_69_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man with short brown hair, wearing a white t-shirt and a dark gray hoodie.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man with short brown hair, wearing a white t-shirt under a dark gray hoodie, which matches the description. The background is plain and does not introduce conflicting elements. The subject's attire and appearance are consistent with the specified 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a young man with short brown hair, wearing a white t-shirt and a dark gray hoodie. which matches the description provided The individual is consistent, does not distract any elements.\"\n The video's attire and hairstyle are consistent with the given 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3c8e13c60ac01f92e5a23e45a9ddc2e852c5f8a5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c2e204898815a5bc762fa00a57fec84e8d2bb3bc5a9e7f013d92f76af5b97aa5
+size 144062
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..5341922313de8944ec2237c4e27636e0e7369196
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a snowboarder's journey down a snowy mountain slope. The snowboarder, dressed in a vibrant green jacket and a black helmet, is seen in three different positions. In the first frame, the snowboarder is at the top of the slope, preparing to descend. The second frame shows the snowboarder in motion, carving through the snow with their snowboard. The third frame captures the snowboarder at the bottom of the slope, having completed their run. The backdrop of the video is a breathtaking mountain landscape, with snow-covered trees and a clear blue sky. The snowboarder's journey is set against this stunning natural scenery, creating a dynamic and exhilarating visual experience."
+    ],
+    "video_ids": [
+        "5LMjGgrl8Oc_0_0to160"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A snowboarder dressed in a vibrant green jacket and a black helmet.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a snowboarder wearing a vibrant green jacket and a black helmet, which matches the description. The person is also equipped with snowboarding gear, including goggles and gloves, and is on a snowy slope, consistent with the activity. The core description is accurately fulfilled without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a personboarder wearing a vibrant green jacket and a black helmet, which matches the description. The snow is also equipped with aboarding gear, including goggles, a, and is actively a snowy slope, which with the ' of The presence elements of accurately represented.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d8c5c4e6dcf4cd5c7c5e8f26f8ed701f6e054fc0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6bcf90aef42508882ad301b5d2f9f83efa3d86aaf0f29dc0a42606618f55bd7
+size 301243
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..8abce0b841aa7b22e468e3fa6a242b64b7a28a6b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man in a patterned shirt is engaged in a conversation with another man in a white shirt. The man in the patterned shirt is gesturing with his hands as he speaks, indicating an animated discussion. The setting appears to be an outdoor patio or garden, with palm trees visible in the background. The lighting suggests it's daytime. The video has a casual, candid feel to it, capturing a moment of interaction between the two men."
+    ],
+    "video_ids": [
+        "21tcVM5tob8_16_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men - one in a patterned shirt, the other in a white shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men engaged in conversation. One man is wearing a patterned shirt with tropical designs, and the other is wearing a white shirt. These details match the 'Object(s)' condition described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two men, in a. One man is wearing a patterned shirt with a designs, and the other is wearing a white shirt. The details match the descriptionObject(s)' condition provided in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3839a4334c119dd7fe3f3bdc07d6460ee84f3830
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1168-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d40754c2e7a62fac48a13f42370885409d24e5119f03fe20aa821106694cb735
+size 161895
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..d7f74059214f4dc58c2baa71ce647019713f164e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a Volkswagen car's front wheel and fender. The wheel is black with red accents and features the Volkswagen logo in the center. The fender is white with black accents and has a distinctive design with air vents. The car appears to be a modern, sporty model. The style of the video is a dynamic, close-up shot that focuses on the details of the car's design and features. The video captures the car's sleek and stylish appearance, highlighting its sporty and modern design."
+    ],
+    "video_ids": [
+        "ZNSagaCFDwI_9_0to142"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Black wheel with red accents and Volkswagen logo, white fender with black accents and air vents.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a black wheel with red accents and the Volkswagen logo at the center, as well as a white fender with black accents and air vents. These elements are prominently featured and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a black wheel with red accents and a Volkswagen logo, the center. which well as a white fender with black accents and air vents. The elements match consistent featured and match the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ba84c257f8ed67f80b58139250b839615059a47e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e403200a44b8687781fdc2fe73175b080762d1e40d3fc47bd6fe1eb366a0202e
+size 148591
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..91cebdc4bcbc3284725e9266a7ad1bf9d95ec95e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a thrilling off-road adventure in a forested area. Two all-terrain vehicles (ATVs) are seen driving down a muddy trail, kicking up dust and leaving a trail of tracks behind them. The first ATV, leading the way, is slightly ahead of the second one. The trail they're on is surrounded by lush greenery, with trees and bushes lining the sides. A small stream runs parallel to the trail, adding a serene element to the otherwise adventurous scene. The video is shot from a high angle, providing a bird's eye view of the action. The style of the video is dynamic and adventurous, capturing the excitement of off-roading in a natural setting."
+    ],
+    "video_ids": [
+        "SRrD13p9kGA_29_0to184"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two all-terrain vehicles (ATVs)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two all-terrain vehicles (ATVs) in motion on a dirt road through a forested area. One ATV is closer to the camera, and another is further ahead, both consistent with the 'Object(s)' condition of the prompt.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts two all-terrain vehicles (ATVs) navigating a through a muddy path surrounded a forest. area. The ATV is in to the foreground, while the is further ahead, both consistent with the descriptionObject(s)' condition described the description.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..208cd0ca1b4116ce037cc6c9134d5ce6ef6700f1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b2691c6d958b10225579cda292eddb8a4b6490294936f624cb3913ea194d7a4
+size 263883
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..0353ada521467840abd25806a2512241f366b9b6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a charming scene set in an outdoor park-like environment. A dog, possibly a Belgian Malinois, is the central figure, standing between two colorful agility hurdles painted in alternating blue and yellow stripes. The dog is holding a basket filled with yellow balls, likely tennis balls, in its mouth. The background features lush green trees and a blurred sign that reads \"DOG MASTY,\" suggesting the location might be a dog training facility or park. The dog appears to be in motion, moving forward slightly while maintaining its grip on the basket. The camera remains stationary throughout the sequence, focusing on the dog's determined expression and the vibrant colors of the hurdles against the natural backdrop. The overall atmosphere is playful and energetic, highlighting the dog's agility and the joyous activity taking place."
+    ],
+    "video_ids": [
+        "90607a738ecc22bbc0f33fa1b19d864f5a3231bd7437820e470a422b700f547e"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A dog (possibly a Belgian Malinois), two agility hurdles painted in alternating blue and yellow stripes, a basket filled with yellow balls.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dog, consistent with the appearance of a Belgian Malinois, carrying a basket filled with yellow balls. It is positioned in front of two agility hurdles painted with alternating blue and yellow stripes. All core elements described are present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dog, which with the description of a Belgian Malinois, running a basket filled with yellow balls. The is running between front of two agility hurdles painted in alternating blue and yellow stripes. The elements elements of in present, accurately represented in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8f2bc502c4af75c9085bb9142646a129152e0734
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f9fdd7308e67ecad4889a83dc6fe4c7bd039bad338d2519399e3c10b3440c004
+size 236099
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..16eb11afde2375f1a5377afad7c9c419f15aede3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a cooking tutorial, showcasing the preparation of a dish. The style is a close-up, high-resolution shot of the ingredients and the process. The ingredients include fresh vegetables such as green peppers, cucumbers, and leafy greens, as well as eggs and rice. The vegetables are vibrant and fresh, indicating that they are being used in a healthy and nutritious meal. The eggs are cracked open and ready to be cooked, and the rice is in a bowl, ready to be cooked as well. The overall atmosphere of the video is clean and organized, with the ingredients neatly arranged on a white countertop. The focus is on the ingredients and the process of cooking, rather than the chef or the kitchen. The video is likely to be informative and educational, providing viewers with a step-by-step guide to preparing a healthy and delicious meal."
+    ],
+    "video_ids": [
+        "iBD0s5JD6fY_6_67to198"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Fresh vegetables (green peppers, cucumbers, leafy greens), eggs, rice? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows fresh vegetables including leafy greens, cucumbers, and red chilies (which can be considered a type of green pepper in some contexts). Eggs are visible in a bowl, and rice is present in a glass bowl with a spoon. All core elements mentioned in the description are clearly visible.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful fresh vegetables ( greeny greens and greenumbers, and green bellilies,which can be considered a type of green pepper). this contexts). It and also, the bowl, and there is also in another separate bowl. a spoon. The these elements of in the description are present visible in}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..46cd4af3c84961df3b625ebeab55aca4a502ee96
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c278b9abce15b810fac066be86d620d232a8c03c98daeafe8418eed01767ca2
+size 132239
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..f6b7d2a43a6ff4d997bff61a02bd548997cba65c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in front of a vibrant orange background with a pineapple design. He is dressed in a gray blazer and a striped shirt, and he is smiling. The man is the main subject of the video, and he is positioned centrally in the frame. The background is bright and colorful, providing a stark contrast to the man's attire. The video appears to be a still image, as there is no indication of movement or action. The overall style of the video is simple and straightforward, with a focus on the man and the colorful background."
+    ],
+    "video_ids": [
+        "YmRFUnK87As_36_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a gray blazer, a striped shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a gray blazer over a striped shirt, which matches the core description. Additional elements like the background and accessories do not contradict the specified objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a gray blazer and a striped shirt, which matches the description description. The elements such the pineapple with the do not contradict the main objects.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f1531a53bfcfd623b29366d4d3c5c97fd84b4bfe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7c2d30d1d801d0ffaa6065a2ea942c243441d483da67a8a12435263ffbbbbf4a
+size 143880
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..7eaab92328d210ac7c52540b52dd16e22177d0a7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures an intense moment in a basketball game. Two players are in mid-air, both reaching for the ball. The player in the foreground, wearing a blue jersey with the number 10, is slightly ahead of the player in the background, who is wearing a white jersey with the number 23. The player in the white jersey is attempting to block the shot, while the player in the blue jersey is trying to score. The background shows a crowd of spectators watching the game, and the basketball hoop is visible in the top left corner of the frame. The style of the video is dynamic and action-packed, capturing the excitement and competition of the sport."
+    ],
+    "video_ids": [
+        "IYfot34F28E_52_21to157"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two players in mid-air, one in a blue jersey with number 10 and the other in a white jersey with number 23. A crowd of spectators and a basketball hoop are also present.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two basketball players in mid-air: one in a blue jersey with number 10 (Memphis) and the other in a white jersey with number 23 (Cavaliers). A basketball hoop is visible at the top left, and a crowd of spectators can be seen in the background. All elements described in the condition are present and consistent with the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts two players players in mid-air, one in a blue jersey with the 10 andSphis) and the other in a white jersey with number 23 (Memeltiers). There crowd hoop is visible in the top of, and a crowd of spectators is be seen in the background. The elements match in the condition are present in match with the video depicted}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3a7145de9a7e234b61c1240047665df5fbaeb921
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c86e4df0d767dc15e1e0650ea550c454350821c38ad076054a9339acb7d06571
+size 260579
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..48a893ce33cfc3b1b44a851a559197126b7ae18c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a gym, performing an exercise on a bench. He is wearing a black shirt and a headband, and he has a ring on his finger. The gym is equipped with various weights and exercise machines. The man starts by lying down on the bench, then he lifts his legs up and holds them in the air. He repeats this exercise several times, demonstrating his strength and endurance. The gym appears to be well-maintained and clean, with a focus on fitness and health. The man's determination and focus are evident in his actions, suggesting that he is dedicated to his workout routine."
+    ],
+    "video_ids": [
+        "aOBiwCeJpZs_36_0to116"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black shirt, wearing a headband, with a ring on his finger, using a bench for exercise.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black shirt and a headband, with a ring visible on his finger. He is positioned on a bench, performing an exercise that involves bending his torso forward with his hands clasped behind his back. The setting appears to be a gym, and the man's attire and actions align with the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a black shirt and a headband, performing a ring on on his finger, He is using on a bench, performing an exercise that involves lifting his legs upwards and his arms behindped behind his head. The setting appears to be a gym, which the man is attire and the align with the description of.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a1fd2ba2c07aecc7107f6e5dd057d87e2b62b7c3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:71c2ee8f13c25f43da1a08a8b95c84d6c1bcf9d1dbe7065b5c0a88b7dc60e416
+size 189729
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..906da83c17c8dbb1c3069eda0c2ee5285fa77f53
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment on a football field. The main focus is a player wearing a red and white uniform with the number 11 prominently displayed. He is wearing a red helmet and a white face mask, and his mouth is open as if he is shouting or calling out a play. The player is in motion, running towards the camera, which adds a sense of urgency and excitement to the scene. The background is filled with other players, some of whom are wearing helmets and pads, indicating that they are also part of the game. The overall style of the video is action-packed and energetic, capturing the intensity and passion of the sport."
+    ],
+    "video_ids": [
+        "3f_KEZLgG7s_7_149to270"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player wearing a red and white uniform with the number 11, other players in helmets and pads.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a red and white uniform with the number 11, wearing a helmet and pads, consistent with the description. Other players in helmets and pads are visible in the blurred background, fulfilling the 'other players' condition. There are no elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player wearing a red and white uniform with the number 11, which a helmet and pads, which with the description. The players in similar and pads are also in the background background, fulfilling the 'other players' condition. The are no elements in contradict the given description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..aff7ec495dcf967efcc00bee0529b6835b0e0c0d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1169-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8077a4ae1b507551058381ca648449a7a784e9dee786660bca0b0f46346da55e
+size 295354
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..9d77253a9111d73a6661ac49acfc76b39f16bc56
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the iconic One World Trade Center in New York City, showcasing its modern architecture and towering presence. The first frame presents a wide shot of the building, its sleek, white facade gleaming under the clear blue sky. The second frame zooms in, revealing the intricate details of the building's design, including the large, white wings that extend from its sides. The third frame offers a ground-level perspective, with people walking around the base of the building, providing a sense of scale and the bustling activity around this landmark. The video style is a blend of architectural photography and cityscape documentation, capturing the essence of New York City's skyline and the impact of the One World Trade Center."
+    ],
+    "video_ids": [
+        "3mRjJkUhxZo_22_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: One World Trade Center, people walking around the base of the building.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows One World Trade Center in the background, towering over the Oculus structure. People are visible walking around the base of the Oculus, which is the central focus of the scene. The description is accurately reflected in the video, with no significant contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows One World Trade Center, both background, fulfilling over the scene building. The are visible walking around the base of the building, which is the lower building of the video. The video of largely represented in the video content with the contradictions contradictions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8ce5e0a99c87bd22bb648d0a66514501970394a2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd96aa9168a48bce72c8c175e266940640a636eb502fa7a5eafe48905e5c9d5c
+size 366577
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..0bd1748a0efde2488452931a19799fbebb87f592
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated scene featuring five children gathered around a baby in a crib. The children are diverse in appearance, with different hair colors and styles, and they are all smiling and looking at the baby with affection. The crib is white with a red and yellow striped mattress, and the baby is lying down, looking peaceful. The room has a pink and white striped wallpaper, and there is a window with a white frame in the background. The overall style of the video is colorful and cheerful, with a focus on the children's expressions and the baby's calm demeanor."
+    ],
+    "video_ids": [
+        "HN5ibJuL9xQ_9_0to102"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Five children and a baby in a white crib with a red and yellow striped mattress. The children are diverse in appearance, with different hair colors and styles, and they are all smiling and looking at the baby with affection.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows five children surrounding a white crib with a red and yellow striped mattress, and a baby inside. The children are diverse in appearance, with different hair colors and styles, and they are all smiling and looking at the baby with affection. The scene matches the description without significant contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts five children and a baby crib with a red and yellow striped mattress. where a baby lying the The children are diverse in appearance, with different hair colors and styles, and they are all smiling and looking at the baby with affection. The scene matches the description provided any contradictions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..59a78adad2c430c44cc34d39d0b4231d107cd0d9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4c6969293766c066a1c284760a8b29ba90ede7179e96c2ad500bba6a15667fc6
+size 175870
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..e60fee48749a9ccfe2fd6ba66dfe0baaecac27c2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit and tie, standing in front of a backdrop that includes a city skyline and the European Union flag. The man appears to be speaking or presenting, as suggested by his open mouth and gesturing hand. The style of the video is professional and formal, likely intended for a news or business broadcast. The man's attire and the backdrop suggest a serious and authoritative tone. The backdrop is colorful and dynamic, with the city skyline and the European Union flag adding a sense of location and context to the video. The overall impression is of a well-produced and polished video, designed to convey information or news in a clear and professional manner."
+    ],
+    "video_ids": [
+        "pzj6MbDVaXk_3_25to157"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, gesturing and possibly speaking or presenting.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed in a suit and tie, standing in front of a backdrop with European Union flags and an Apple logo. He appears to be gesturing with his hands and his mouth is moving, suggesting he is speaking or presenting. These elements align with the described 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man in in a suit and tie, which in front of a city that a Union flags and a urban logo, He is to be gesturing with his hands, possibly mouth is open, suggesting he is speaking or presenting. The elements align with the description 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5fdfbfc084d1de22f9b7e48da2283dd9edade173
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41fe69138d5afd75a82351f89685bc77e1471d723d34991d27b1fd802255ac4d
+size 190833
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..ad82653ee2fde57b2e0380bcccd40205ec99bb9f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman standing in front of a modern building with large windows. She is wearing a white sweater and a necklace. The building has a curved design and is surrounded by a well-maintained lawn. The woman appears to be speaking or presenting, as she is looking directly at the camera. The overall style of the video is professional and polished, with a focus on the woman and the building. The lighting is bright and even, highlighting the woman's features and the building's architecture. The video does not contain any text or additional graphics."
+    ],
+    "video_ids": [
+        "PMXVK3BS0r8_6_0to185"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a white sweater, a necklace, a building with large windows and a curved design, a well-maintained lawn.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a woman wearing a white sweater and a necklace, standing in front of a building with large windows and a curved design, with a well-maintained lawn visible in the background. All specified elements are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a woman wearing a white sweater and a necklace. standing in front of a building with large windows and a curved design. with a well-maintained lawn in in the background. The the elements are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..38de492d5de00ef23aa750389fad8ef7820dde4b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:348b8f28ebbe12cad06ded5b895d6062f3670293a3e76c4d789970f3b34c3aab
+size 73715
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..b5ccc024ed67c37a46339f11377bf7a6fc1b33a1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video opens with a dramatic scene set in an open field under a cloudy sky. A large, green, tarp-covered structure dominates the frame, its surface marked by numerous cracks and holes, suggesting it has been subjected to some form of impact or damage. Perched atop the tarp is a small, dark-colored animal, possibly a weasel or similar creature, which appears to be observing something intently. Suddenly, a red ball is launched towards the tarp, striking it with force. The impact causes a burst of white powder or dust to erupt from the point of contact, creating a stark contrast against the green tarp. The camera remains stationary throughout, capturing the entire sequence of events without any noticeable movement or change in angle.  In the bottom right corner of the video, there is a smaller inset image showing four individuals standing together in the same field. They appear to be engaged in conversation or reacting to the event unfolding on the larger screen. Their expressions and body language suggest amusement or"
+    ],
+    "video_ids": [
+        "8fc1eb69b806f22f24f6b3b74e605dfa19a259edbcefb1a63973bf2c97b2fc7f"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Large green tarp-covered structure, small dark-colored animal, red ball.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a large green tarp-covered structure with visible cracks, a small dark-colored animal (resembling a dinosaur toy) perched on top, and a red ball positioned near the animal. These core elements are clearly present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a large green tarp-covered structure, a folds, a small dark-colored animal perpossiblysembling a bear)) onched on top, and a red ball that near the structure. The elements elements match present present, match the description.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5f91af9b20d90c23d2e7e12c54ff387d4ce9117a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b7615cf722348f23b0bd8b398ac6df9acae49dea298e35223952ccc4dcc2979e
+size 132767
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..b7460761a95e3e275e399b3382cf42b78a166b91
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a brown hat and glasses, wearing a brown leather jacket, standing in front of a store window. He is holding a red microphone with a white \"A\" on it. The man appears to be speaking or giving an interview. The store window behind him displays various items, including toys and books. The scene suggests an urban setting, possibly a shopping area or a street with commercial establishments. The man's attire and the microphone indicate that he might be a reporter or a public speaker. The overall style of the video is candid and seems to capture a moment in the man's day, possibly during a public event or a news report."
+    ],
+    "video_ids": [
+        "BW0xPlFdWAY_0_0to172"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a brown hat and glasses, wearing a brown leather jacket, holding a red microphone with a white 'A' on it. Store window displaying various items, including toys and books.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a brown hat and glasses, dressed in a brown leather jacket, holding a red microphone with a white 'A' on it. Behind him, there is a store window displaying various items, including toys and books, which matches the description. The scene is consistent with the given object conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a brown hat and glasses, dressed in a brown leather jacket, holding a red microphone with a white 'A' on it. The him, there is a store window displaying various items, including toys and books, which align the description provided The presence is consistent with the provided description conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4fb42168cef0093452555d6e090ce0596fd0fbc0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea37234f3ad247062604df272f4a5d63071d3d6c7af6a104624ed2d344f842a9
+size 182773
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..8a741e5fe7a0237e2e07fd0b7602352f4767215a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a drone shot of a mountainous landscape, showcasing a winding dirt road that cuts through the dense forest. The road is surrounded by lush greenery and towering mountains, creating a sense of isolation and tranquility. The perspective of the drone provides a bird's eye view of the landscape, highlighting the rugged terrain and the winding path of the road. The video captures the beauty of nature and the solitude of the road, making it a perfect representation of a serene and untouched wilderness."
+    ],
+    "video_ids": [
+        "X4wsDhTkB9A_13_0to126"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Winding dirt road, lush greenery, towering mountains? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by prominently featuring a winding dirt road that snakes through the landscape, lush greenery covering the hillsides, and towering mountains in the background. These elements are clearly visible and align with the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting featuring a winding dirt road that me through lush landscape. surrounded greenery covering the valleyides, and towering mountains in the background. The elements are clearly visible and dominate with the description provided.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f95e5ec20f8282b0725613eaf6794b0d77bd4362
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9d450f2202a2a0dda917036de4b9a73a48ffb22e148f8f5d4268541722e87f25
+size 188291
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..20be6abbfa17d8ad1c428932ad53c4376da8f960
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in front of a large building with a tree in the background. He is dressed in a suit and tie, and appears to be speaking or presenting. The man is holding a piece of paper, which he seems to be reading from. The style of the video is straightforward and professional, with a focus on the man and his message. The setting suggests an urban environment, possibly a government or corporate building. The tree in the background adds a touch of nature to the otherwise urban scene. The man's attire and the paper he is holding indicate that he may be a public speaker or a professional giving a presentation. The video does not contain any other significant objects or actions."
+    ],
+    "video_ids": [
+        "cRYORSjWmBE_2_0to151"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a suit and tie, holding a piece of paper.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed in a suit and tie, holding a piece of paper, which matches the described condition. The background and additional elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed in a suit and tie, holding a piece of paper. which align the description '. The setting includes additional elements, not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..31dc7194c7b4e686d7d5be164d7848084d08a2ce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-117-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9729663a9376a84e1d02b04e6e02db4d35725d4b1147cb5aa9ef781d90b2014f
+size 165743
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..89285fe6c5d4cf92dd48447a7df7bda6609750f6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "Three men in a wine cellar, each holding a glass of wine, discussing and comparing their drinks. The cellar is filled with a variety of wine bottles, and the men are casually dressed, suggesting a relaxed and friendly atmosphere. The lighting is warm and inviting, highlighting the rich colors of the wine and the wood of the cellar. The men are engaged in conversation, with one man gesturing as he speaks, indicating an animated discussion. The overall style of the video is casual and social, capturing a moment of shared enjoyment and camaraderie among friends."
+    ],
+    "video_ids": [
+        "bi3v435Fb-k_13_23to159"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three men, each holding a glass of wine.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three men, each holding a glass of what appears to be dark beer or wine, and they are engaged in tasting or discussing the beverage. The setting and their actions align with the core description, even though the beverage may not be wine, the condition is met as the men are holding glasses of a drink.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three men standing each holding a glass of wine appears to be wine red, wine, standing they are to in a or drinking the beverage. The setting is the actions align with the description description of fulfilling though the beverage is not be wine as it condition of met in they men are holding glasses and a liquid.\"\n}<|im_end|>\nGuidId",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cbbc681b8efe501f6b4ddc028377632a7b9b21d6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d0e474eca82c74831b8ac6b5fbb4f1250f36295269eb03289bfb3f239df076bd
+size 204108
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..8a44a541e449ff457100b1bbe415a7c82e419ba4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of two glass containers filled with a meal. The meal consists of sliced chicken breast, brown rice, and a variety of vegetables including corn, black beans, and avocado. The containers are placed on a gray surface, and the meal is garnished with chopped cilantro. The style of the video is a simple, straightforward food presentation, with no additional context or setting provided. The focus is solely on the meal, showcasing its ingredients and presentation."
+    ],
+    "video_ids": [
+        "2ILZfU8NtZM_11_0to126"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two glass containers, sliced chicken breast, brown rice, corn, black beans, avocado, chopped cilantro.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two glass containers filled with sliced chicken breast, brown rice, corn, black beans, avocado, and chopped cilantro. All the specified elements are present and accurately represented without any contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows shows two glass containers with with sliced chicken breast, brown rice, corn, black beans, avocado, and chopped cilantro. The the elements elements are present, match depicted in any contradictions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ea8e2fc2f36e448bd6bd7a0fb94c16677c43e7f9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:576e4d42a4abeea05cc0ca026c0abdbf56106b41c92724e64a9b952dc9c35f37
+size 79298
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..3cd067bcc845a30eed4093c3b23c500549ca7fa2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a white car with black rims and tires, parked on a clear day. The car is positioned on a glass platform, which reflects the car's shiny exterior. The platform is set against a blue background, which contrasts with the car's white color. The car is stationary, and there are no people or other objects in the frame. The style of the video is a straightforward, clear shot of the car, with no additional action or movement. The focus is solely on the car and its details, such as the rims and tires. The lighting is bright and even, highlighting the car's features without creating harsh shadows. The overall impression is of a clean, well-maintained vehicle, presented in a professional and polished manner."
+    ],
+    "video_ids": [
+        "UPKjXZf3jPE_11_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: White car with black rims and tires? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white car with black rims and tires, which matches the description. The camera pans across the side of the car, clearly displaying the black wheels and the white body. There are no conflicting elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a white car with black rims and tires, which matches the description provided The car angle around the car of the car, highlighting showing the white rims and the white body of The are no additional elements in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..be7651145d3584ff5ee546e030a78180bacecaa5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54ee03cbe4ee6dcd3323b31a15302f221544a607328c8d8fdfb14e325fa072b5
+size 137207
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..ae9e41e48569965d68432bf0e1634bb699c21985
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a car engine being worked on. The engine is open and exposed, revealing various components such as the air filter, spark plugs, and fuel lines. The engine is silver and black, and it appears to be in the process of being serviced or repaired. The style of the video is informative, with a focus on the details of the engine and the work being done. The video is likely intended for educational purposes, such as teaching viewers about car maintenance or engine repair."
+    ],
+    "video_ids": [
+        "bvez36pq1oc_23_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Car engine, air filter, spark plugs, fuel lines? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a car engine with visible components such as the air filter housing, spark plug wires (implied by the engine setup), and fuel lines (indicated by the various hoses and connections). Although spark plugs themselves are not individually visible, they are typically part of the engine assembly shown, and the video largely matches the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a car engine, various components such as the air filter,, spark plugs wires,whichplied by the presence's), and fuel lines (imicated by the blue tubes and tubes). The the plugs are are not directly visible, the are part part of the engine setup and in and the presence does matches the description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ce0aec5dd54eebe92f1c70c4c4cd95ff92a84b04
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c606e32d04ab0203b678d0a13150834559799722e8cce3e5daffa80fcaf16f59
+size 151889
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..fedb5a6558eab28fea8ad640c74fbd00bec604ad
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a close-up view of a green Aston Martin car, focusing on its rear end. The car is equipped with a black spoiler and a black exhaust pipe. The car's black rims and black tires are also visible. The car is parked in a garage with a white floor and white walls. The garage has a high ceiling and is well-lit. The car is the main object in the video, and the camera angle is from a low perspective, emphasizing the car's design and details. The video is likely a promotional or review video for the Aston Martin car."
+    ],
+    "video_ids": [
+        "Vi7n8-LIe4I_49_0to143"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A green Aston Martin car with a black spoiler, black exhaust pipe, black rims, and black tires.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a green Aston Martin with a black spoiler, black exhaust pipe (visible as part of the rear diffuser), black rims with silver accents, and black tires. The car's design and color scheme match the description, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a green car Martin car a black spoiler, black exhaust pipe,visible in part of the rear designuser), black rims, black accents, and black tires. The car's design and color scheme match the description provided fulfilling there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fe9e9dc21668d1e5a2241155b15654a963c0502a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc5f863bc08ed487c0ffa9efec51ed86b18fcf5da6a92810a4904eb9e5152700
+size 76292
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..7b8865105ab4500b5545b78feac879fc93ef3af2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a blue animated character with a helmet, who is holding a large roasted chicken. The character is excited and surprised by the size of the chicken, as indicated by its wide-open mouth and eyes. The character is standing on a boat, which is floating on a body of water. The background is a clear blue sky, suggesting a calm and sunny day. The style of the video is cartoonish and colorful, with a focus on the character's exaggerated facial expressions and the vibrant colors of the chicken and the water. The overall mood of the video is playful and lighthearted."
+    ],
+    "video_ids": [
+        "S7MlcDHlkpU_40_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue animated character with a helmet holding a large roasted chicken.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a blue animated character wearing a helmet, who is holding a large roasted chicken. This matches the core description. Additional elements like the jet ski and shark do not contradict the main object condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a blue animated character wearing a helmet, holding is holding a large roasted chicken. The matches the description description provided The elements such the backgroundpack and the in not contradict the main object condition but}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3e0323dcfc048e01091478e3621dd716ed224dee
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6aa15e5b258388e6e44556c4d888edf11d54c83063a63481f114f9fc63ef0c69
+size 243233
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..47b4b5829f3166674f4d1b103d7890aa14191af1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a plate of food on a table. The plate is white with a green and yellow pattern around the edge. The food consists of a piece of meat, possibly chicken, topped with shredded vegetables and garnished with a sprig of parsley. The tablecloth is black, and there is a fork to the right of the plate. The style of the video is a simple, straightforward food presentation, likely intended to showcase the dish's presentation and ingredients. The focus is on the plate and the food, with no additional context or background provided."
+    ],
+    "video_ids": [
+        "RVDXKYWwtzo_6_19to174"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white plate with a green and yellow edge, a piece of meat (possibly chicken), shredded vegetables, and a sprig of parsley. A black fork is to the right of the plate.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white plate with a green and yellow patterned edge, containing what appears to be a piece of meat (possibly chicken), shredded vegetables (including red bell peppers and possibly ginger or daikon), and a sprig of parsley on top. A fork is visible to the right of the plate, although it is silver, not black. The presence of a silver fork instead of a black one is a minor deviation but does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a white plate with a green and yellow edgeed edge. which shredded appears to be shredded piece of meat (possibly chicken), shredded vegetables,yellow red and peppers and possibly z), anotherikon rad and a sprig of parsley. top. To black is visible to the right of the plate, matching it is not rather not black. The presence of the fork fork does of a black one does the minor discrepancy from does not significantly the core description.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dfccfba2d54761adb583e0b96755618f2adfa724
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5bd3656714e52e562f5397a2eb54e65c244d5b6b790983bdc8375de20ffee81
+size 86923
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..00bf12a728b6323b789cae943e554c64488121d9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of two boats sailing on a clear blue body of water. The boats, one larger and one smaller, are moving in the same direction, leaving a trail of ripples in their wake. The larger boat is closer to the camera, while the smaller boat is further away, both appearing to be moving towards the right side of the frame. The water is a vibrant blue, reflecting the clear sky above. In the distance, a small island can be seen, adding a touch of land to the otherwise expansive water body. The overall style of the video is calm and peaceful, capturing the beauty of a day out on the water."
+    ],
+    "video_ids": [
+        "AG-TYpMi6Ek_27_743to937"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two boats (one larger, one smaller), water, sky, island? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two boats \u2014 one larger sailboat and one smaller motorboat \u2014 on turquoise water, with a visible island in the background and a clear sky above. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts two boats, one larger andboat and one smaller motorboat \u2014 on water water. with a visible island in the background. a clear sky above. The the elements of in present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bd49c5c7c15f33233a6e3f8e2e9deca0ee45e338
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1170-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a7e7e96080ba6e6253a617c19bc45678013d727bd68627c440b30308b152297
+size 467341
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..bd7e1345af4b4e4b82afd645c37a24d1af179a0d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person holding a small white container with a piece of food inside. The food appears to be a fried item, possibly a pastry or a dumpling, and it is accompanied by a green sauce or condiment. The person is standing in an outdoor setting, possibly a street or a public area, as there are other people and objects in the background. The style of the video is casual and informal, with a focus on the food item being held. The lighting is natural and bright, suggesting it might be daytime. The video does not contain any text or additional graphics."
+    ],
+    "video_ids": [
+        "Pq1-hoGxJ4c_42_0to130"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person holding a small white container with a fried food item (possibly a pastry or dumpling) and a green sauce or condiment.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person's hand holding a small white container that contains a fried food item, which appears to be a pastry or dumpling, along with a green sauce or condiment. This matches the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person holding hand holding a small white container with contains a fried food item, which appears to be a pastry or dumpling, and with a green sauce or condiment. The matches the description provided. the questionObject(s)' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e3d733cd9c3b362e88f29553d22d5b7bb56ae016
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f186d064b7003c269ab6f54f2f34db7a25d1dcdbd10afe2316d2cc41444a765
+size 143909
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..c809b0288549790baf4615a76ce596e0be9125ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are standing on a stage with a backdrop that features the word \"REAL\". The man is dressed in a blue suit and tie, while the woman is wearing a red dress. They are both holding cards in their hands. In the foreground, there is a glass bowl filled with pink slips of paper. The stage appears to be set for a game or a contest, and the atmosphere seems to be lively and engaging."
+    ],
+    "video_ids": [
+        "MUz6WSEh73w_10_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a woman, cards, a glass bowl, pink slips of paper.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a woman on a set, with a glass bowl filled with pink slips of paper. The woman is seen pulling out and holding a pink slip of paper, which matches the described objects. The presence of additional elements (like the set design) does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a woman standing a stage, each the glass bowl containing with pink slips of paper. Both man is holding holding a a holding a slip slip of paper, which align the description elements. The man of cards elements likelike the set background and does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f51f45213f863d76cac60bc43c8ab8c59ccbcc76
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:82de0d0b3ee847cc846574a9d0458be7a5aa12f7451c811afb62bc9a77bf040b
+size 186465
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..2438d2272c46493b25aaa26bd66072e114a1c70a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the construction of a large building in a city. The first frame shows the building in its early stages, with the foundation laid and the skeleton of the structure beginning to take shape. The second frame shows the building further along in the construction process, with the walls and floors starting to be built. The third frame shows the building nearing completion, with the exterior walls and windows installed. The style of the video is a time-lapse, showing the progression of the construction over time. The video provides a unique perspective on the construction process, allowing viewers to see the transformation of the building from start to finish."
+    ],
+    "video_ids": [
+        "4eMHFZ62_QI_24_0to189"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Building under construction? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a building under construction, identifiable by the exposed brickwork, stacked bricks on the roof, and the presence of a large crane actively engaged in construction work. The surrounding area includes construction materials and partially completed structures, all consistent with an ongoing building project.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts depicts a building that construction, as by the unfinished structuralwork and incomplete construction, the roof, and the unfinished of construction crane crane in lifting in lifting activities. The structure area, other materials and equipment completed structures, which of with a ongoing building project.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5c8ebd77ca4f25cf55259e4d309a1f9dc47ee249
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e2cc8f9ca4a38594b0e76d1c6dce38d206344bf8247bb806769bfbf0f4dd97ae
+size 171301
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..a948d5aed53a89e325b5071caf8f8e0092c93e13
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young woman with long blonde hair is seen standing in front of a red car. She is wearing a red jacket and appears to be in a state of surprise or shock. Her hands are raised slightly, and her mouth is open as if she is about to speak. The car is parked on a grassy area, and there is a house in the background. The woman's expression and body language suggest that she has just witnessed something unexpected or startling. The overall style of the video is realistic, with natural lighting and colors."
+    ],
+    "video_ids": [
+        "Um0ZXhrQUJ4_6_0to159"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman with long blonde hair, wearing a red jacket, and a red car.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman with long blonde hair wearing a red top (which can be interpreted as a red jacket or sweatshirt), standing in front of a red car. The core elements described in the condition are present and accurately depicted, even though the background includes additional elements like a house and trees, which do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young woman with long blonde hair wearing a red jacket,which could be interpreted as a red jacket). hoodieshirt). standing in front of a red car. The setting elements of in the question are present and match represented in with though the exact includes a elements like a house and green, which do not contradict the main.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..678bed9102a1501dcb84da22e6066b6a0d2ece2f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c14fd572a6a740629aaa470595cc3fde2c4717e7a87b4e9be612d0a684ebd93
+size 184979
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..9aa23777c1281f62c79b37437e6d71ad24b00569
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a nighttime scene illuminated by a flashlight, creating a stark contrast between light and shadow. The setting appears to be an outdoor area, possibly a backyard or a garden, bordered by a wooden fence on one side. The ground is covered with dry leaves and twigs, suggesting it might be autumn or winter. A black cat with white markings is the focal point of the video. It moves cautiously across the uneven terrain, its eyes reflecting the light from the flashlight. The cat's movements are slow and deliberate, indicating wariness or curiosity about its surroundings. As the camera follows the cat, it pans slightly to the right, revealing more of the dark, leaf-strewn ground and parts of the wooden fence. The overall atmosphere is quiet and tense, enhanced by the limited visibility and the cat's cautious exploration."
+    ],
+    "video_ids": [
+        "806a3c7f6498bfa1be67ea3085acad8cc49af6e4d869ef1cc04b03cd7a9c70db"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black cat with white markings? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black cat with distinct white markings, particularly on its chest and face, which matches the description. The cat is clearly visible under the flashlight in a nighttime setting, and no conflicting elements are present that contradict the description of the cat.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black cat with white white markings, which on its face and p, which align the description of The cat is interacting visible and a lighting, a setting setting, surrounded the other elements are present.\"\n would the core.\"\n the object.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..21e2b1adb531cb3dbb05ad71dd0ce0b072eac60e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3a9d519d7c3830e18599b0ee06e451003fd048ddf95eec612de3094b789d562
+size 206127
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..bb95d32018ca74957daf0bc33ec773a24ff200b2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a football player in a white jersey, standing on the sidelines of a stadium. He is looking off to the side, his expression serious. The stadium is filled with spectators, their faces a blur of anticipation and excitement. The player's jersey is clean and crisp, contrasting with the muddy field beneath his feet. The video is shot in a realistic style, capturing the atmosphere of a live football game. The player's focus and the crowd's energy are the main elements of this video."
+    ],
+    "video_ids": [
+        "h1Pv8Ed3hRo_85_105to235"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Football player in a white jersey, spectators? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video includes a football player wearing a white jersey (visible in frames 0:04-0:09), and spectators are visible in the background of the stadium in those same frames. The presence of additional elements like the player in a blue jersey in the first frames does not contradict the core description, as the condition only requires the presence of the specified object(s).\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a white jersey,with on the 1 and00-0:08) and there are present in the background, the stadium ( frames frames frames. The presence of additional elements like the stadium's a white shirt and the foreground few does not contradict the core description, as the focus only requires the presence of a player objects and)\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7837bb1c32b56e82c97bdebc02046721f13ef777
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:edd28e292e29ec8f8f49ede9def025ba4d2a13879a5b725f708705d502725870
+size 233642
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..247eb657c00f3c8b4a7a12936494c13ffa2f9233
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a journey on a two-lane road through a dry, arid landscape. The road stretches out ahead, with a car driving on the right side of the road. The surrounding environment is characterized by dry grass and sparse vegetation, with a few trees scattered along the roadside. The sky above is clear, suggesting a sunny day. The perspective of the video is from an aerial view, providing a bird's eye view of the road and the surrounding landscape. The car's movement along the road creates a sense of motion and progress. The overall style of the video is realistic, with a focus on the natural environment and the journey itself."
+    ],
+    "video_ids": [
+        "SnMrJa85MyM_8_133to280"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A car driving on the right side of the road.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white car driving on the right side of a two-lane road, which matches the described condition. The car remains consistently on the right side of the road throughout the footage, and no conflicting elements are present that would contradict this observation.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a car car driving on the right side of a two-lane road, which align the description condition. The road is consistently on the right side throughout the road throughout the frames, and there other elements are present.\"\n would contradict the observation.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..43cd086a83b91464589c610c353fe94b5d3466ea
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:304c6185a1ce26ad68cd2f5d67efb16ecdecdc83ae896aff98ff06ca47fa19bf
+size 366640
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..d77aab9fa9fdee997c9030d9ab8772d757afae65
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a cup of coffee on a floral-patterned saucer, placed on a wooden table. The coffee has a frothy top, indicating it is freshly brewed. In the background, there is a plate of sesame-covered bread, suggesting a breakfast or snack setting. The style of the video is simple and straightforward, focusing on the coffee and the bread, with no additional action or movement. The lighting is soft and natural, enhancing the warm and inviting atmosphere of the scene. The video is likely intended to evoke a sense of relaxation and enjoyment, as one might experience while enjoying a cup of coffee and a snack."
+    ],
+    "video_ids": [
+        "HgxwjbGgNMw_2_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A cup of coffee with a frothy top, a plate of sesame-covered bread.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a cup of coffee with a frothy top, placed on a floral-patterned saucer, and a sesame-covered bread (likely a simit or similar bread) is visible next to it. These elements match the description exactly, with no contradictory elements present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a cup of coffee with a frothy top, which on a sa-patterned saucer. and in plate-covered bread inlikely a typeit or a type) on visible in to the on The elements match the description provided, fulfilling no additional or present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5eddf092330b5e9f8f605d49a5ab7e607a73d43e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1171-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11dbd343ffbd4769dc5f90945238939149caf3768caae4dcab066d36f2fd1251
+size 65869
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..83e0e56daf70b1f184ad91cba8c24fd8f5463077
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of two slices of pizza on a plate. The pizza is topped with a pink sauce, green onions, and cheese. The plate is placed on a tiled countertop. The style of the video is a simple, straightforward food presentation. The focus is on the pizza, with no additional elements or distractions. The lighting is bright, highlighting the colors of the toppings and the texture of the crust. The video does not contain any text or narration. The overall impression is of a casual, everyday meal, captured in a clear and uncluttered manner."
+    ],
+    "video_ids": [
+        "1ao4hxGD7Bo_11_0to170"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two slices of pizza, pink sauce, green onions, cheese? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two triangular slices of pizza on a plate, topped with pink sauce (likely a creamy or mayonnaise-based sauce), chopped green onions, and melted cheese. All core elements described are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two slices slices of pizza with a plate. with with pink sauce,likely a sauce or spreadonnaise-based sauce), green green onions, and melted cheese. The the elements of in present, match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7d9531657455f23fa555a3ac3afea2bf5496a354
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2a94cff42e78ca73e982f69415fe4915c03375201640ee061827172a2497e233
+size 107740
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..c8b7e04d4eebb392c05977da02ac76d2d2700e5f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures an underwater scene featuring a vibrant coral reef. The primary focus is on a large octopus, which is perched atop a cluster of yellowish-brown corals. The octopus's skin displays a mottled pattern of orange and brown hues, blending seamlessly with its surroundings. Its tentacles are extended outward, some resting on the coral while others hang loosely, showcasing the creature's flexibility and adaptability to its environment.  As the video progresses, the octopus begins to move slowly across the coral structure. Its movements are deliberate and graceful, with each tentacle carefully maneuvering over the uneven surface. The camera follows the octopus's journey, maintaining a steady focus on the animal and its interaction with the coral. The background remains consistent, with other corals and marine life visible but slightly out of focus, emphasizing the octopus's central role in the scene.  Throughout the video, the lighting is dim, typical of an underwater setting, casting a soft glow on the oct"
+    ],
+    "video_ids": [
+        "182dc25ed8703ebad650023a384f0bf159460a61068c8bfafa4668e90d669868"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Large octopus, yellowish-brown corals, other corals and marine life.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a large octopus with a reddish-orange hue, resting and moving over yellowish-brown coral formations. The background includes other coral structures and faint marine life, aligning with the description. The scene is consistent with an underwater coral reef environment, and no elements contradict the specified objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a large octopus with a yellowish-b and, which on moving among yellowish-brown cor structures. The presence includes other cor structures and marine hints life, whiching with the description of The vibrant is vibrant with the underwater environment reef environment.\"\n fulfilling the elements contradict the given objects.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..51848fac0b285d2b6f2e58f2bc38ba2699fa9fbb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ad1794ee8b36cd6e123df9ce6897e1245318d09c50bcdfc1cb7d680fd963f69b
+size 260832
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..3d9db953a59751b705d4f048c0f0ed1ab4416f88
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a tray of small, round pastries with a yellow filling. The pastries are arranged in a grid pattern, and each one is topped with a sprinkle of nuts. The style of the video is simple and straightforward, focusing on the pastries without any additional context or background. The lighting is bright, highlighting the texture of the pastries and the contrast between the yellow filling and the nuts. The video does not contain any text or additional elements, and the focus is solely on the pastries. The pastries appear to be freshly baked and ready to be served. The video does not contain any action or movement, and the pastries remain stationary throughout the video. The video is likely intended to showcase the pastries for a culinary or food-related context."
+    ],
+    "video_ids": [
+        "wBZN9nDZ6GM_1_0to123"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Small, round pastries with yellow filling, topped with a sprinkle of nuts.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows small, round pastries with a yellow filling, and many are topped with a sprinkle of nuts (almonds), which matches the description. The arrangement and appearance of the pastries are consistent with the specified object condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows small, round pastries with a yellow filling, which they of topped with a sprinkle of nuts.ormonds and which matches the description provided The past and appearance of the pastries are consistent with the given condition condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..846df498650a9ae8c71dc116993d70ea364bb95d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:146f8a287021828756c01f525f1f3b9166755f6020a5e6ad049869e446e2b866
+size 143461
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..c23cf36b94ca871c4d2f921f07934cba109d6898
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman and a young girl are seen sharing a quiet moment together in a cozy living room. The woman, dressed in a casual gray shirt, is seated comfortably in a blue armchair, her attention focused on the book she's reading aloud to the girl. The girl, dressed in a purple dress, is nestled comfortably in the woman's lap, her eyes wide with interest as she listens to the story. The room around them is warm and inviting, with a bookshelf filled with various books and a clock on the wall indicating the passage of time. The scene captures a tender moment of bonding between the two, as they immerse themselves in the world of the book."
+    ],
+    "video_ids": [
+        "YAMKC_VMnIM_35_0to113"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a gray shirt and a young girl in a purple dress? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a gray shirt and a young girl wearing a purple dress, which matches the specified description. The core elements are accurately represented, and additional background elements (like the chair, bookshelf, and clock) do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a woman wearing a gray shirt and a young girl in a purple dress. both matches the description '. The woman elements of present represented, and there elements elements likelike the book and booksshelf, and books) do not contradict the main.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..83819ca0f03f4409efb894ea48dd220b4f6a869f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b099f0b7c6b404844c30a53314c6f7308fb3447fd00bc83bc87eb2f1d0731cdb
+size 95128
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..4e62d58e5359156fb9a27c454a10973fd880d535
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the process of assembling a pulled pork sandwich. In the first frame, a pile of shredded pork is placed on a white plate. The pork is generously covered in a rich, dark barbecue sauce. In the second frame, a slice of white bread is placed on top of the pork, creating a sandwich. The sandwich is then placed on a white plate with a blue floral pattern. In the third frame, the sandwich is ready to be served, with the plate placed on a table. The sandwich appears to be delicious and ready to be enjoyed."
+    ],
+    "video_ids": [
+        "LL6E_TiDItg_69_0to119"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Shredded pork, white bread, barbecue sauce, white plate, blue floral pattern plate, table? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows shredded pork on a white bread bun, topped with barbecue sauce, all placed on a white plate with a blue floral pattern. The setting appears to be a table, and all core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful the pork, a white plate bun, which with barbecue sauce. placed placed on a white plate. a blue floral pattern. The table appears to be a table, which the the elements of in present in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..55aee5f8c76e5bc058d01568d6aa66a0f54c6ae6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87d4575d7276faf54bb7630761ef0876fb06a9278a9b30adf3eda22d2353bc4c
+size 190099
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..a5b95de9224a2eb9a367cce9b34229633572bcf4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a gym, wearing a grey tank top with the word \"REGENICS\" printed on it, and a black baseball cap. He is smiling and looking up, possibly at a camera or someone off-screen. The gym is well-equipped with various exercise machines and weights, and the lighting is bright, illuminating the man and the gym equipment. The style of the video is casual and friendly, with a focus on the man and his surroundings. The gym appears to be a modern and well-maintained facility, with a clean and organized layout. The man's attire and the gym environment suggest that he is engaged in a workout or fitness routine."
+    ],
+    "video_ids": [
+        "4BGkPqo3BtU_15_0to173"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a grey tank top with 'REGENICS' printed on it, and a black baseball cap.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a grey tank top with 'REGENICS' printed on it and a black baseball cap, which matches the core description. The background and other elements do not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a grey tank top with 'REGENICS' printed on it and a black baseball cap. which matches the description description provided The setting appears additional elements in not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2c6e94254f56bc1c309aa9afa451e45d51b746ac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a918684a4cfd4002e8c3a4ff192dc535d685d02cdbea3548d41012997f076097
+size 173342
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..2cdaa5aef7eb772f3c5e0dae17a5e2e3823d23b5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a blue sweater vest and a white shirt, standing in an office-like setting. He is speaking into a microphone with a blue and yellow logo, which is likely a news station's logo. The man appears to be in the middle of an interview or a news report, as suggested by the presence of the microphone. The office setting includes a desk and a window, and there are framed pictures on the wall. The man is the main subject of the video, and his facial expression suggests that he is engaged in a serious conversation or discussion. The style of the video is a standard news report or interview, with a focus on the man and his interaction with the microphone."
+    ],
+    "video_ids": [
+        "HRfOrb-C-Pk_8_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a blue sweater vest and white shirt, speaking into a microphone with a blue and yellow logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue sweater vest over a white shirt, speaking into a microphone that has a blue and yellow logo (WBZ). The core description is accurately represented, and additional elements like the background and on-screen graphics do not contradict this.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue sweater vest and a white shirt, holding into a microphone. has a blue and yellow logo.which).). The setting elements is accurately represented in and there elements like the background and the-screen graphics do not contradict the description}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..950093a36af3ce1e2b6fe0d7fa542a4ee86ce9b8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e993fb951d7576200b7459c40d6eff95bd87a9ffbe3681c7b6a9409411e37ac3
+size 208163
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..c373cd3c935bdc5c07d278201b8e799fecb59d75
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a gorilla in a zoo enclosure. The gorilla is seen in three different positions: standing, sitting, and bending over. The gorilla is surrounded by a naturalistic environment that includes a tree, rocks, and a grassy area. The gorilla is the main focus of the video, and its actions are the primary subject. The style of the video is a straightforward, unedited documentation of the gorilla's behavior in its enclosure. The video does not include any additional elements or effects."
+    ],
+    "video_ids": [
+        "mliim9maTVg_24_128to284"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A gorilla? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a gorilla sitting and interacting with vegetation, which aligns with the 'Object(s)' condition of featuring a gorilla. The background elements, such as trees and rocks, do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two pairilla in on interacting with another, which aligns with the descriptionObject(s)' condition of depicting a gorilla. The presence and, such as the and grass, do not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8cf51d61bbbb66119994a4598cd063637e126209
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1172-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d71a7fa936d8483830b13cde95676797626f4a33345d5f9a515059881cd450c8
+size 194802
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..c5dcc340c85f69b8e584d9b78a8e6cd3d77902b4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a car driving on a winding mountain road. The car, a sleek and modern model, is seen from a low angle, emphasizing its speed and agility as it navigates the sharp turns of the road. The road itself is a marvel of engineering, carved into the side of a steep mountain, with guardrails visible in the distance. The surrounding landscape is rugged and wild, with rocky cliffs and sparse vegetation dotting the landscape. The sky above is a clear blue, with a few clouds scattered across it, adding to the sense of openness and freedom. The overall style of the video is dynamic and adventurous, capturing the thrill of driving on a challenging road in a beautiful and remote location."
+    ],
+    "video_ids": [
+        "g9owvTBTJdQ_67_92to217"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A sleek and modern car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a silver car that appears sleek and modern in design, consistent with the description. The car's smooth lines and contemporary styling match the 'sleek and modern' requirement. The surrounding environment does not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a car car that appears sleek and modern, design, which with the description of The car is design lines and contemporary styling are the 'sleek and modern' criteria. The setting environment, not detr the description,}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fb59f705dd99cc1f5c8adb4bd8c53219708c6fe5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c6225774f0bcfb871be78c0fb45cdbcef309a3290c841b9985186a769fc3396
+size 250709
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..fec72e861d5d95e6c829b54d05460438ed90a26b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a black sports car driving down a tree-lined road. The car is sleek and shiny, with a distinctive front grille and a spoiler on the back. The road is curving to the left, and the car is moving at a moderate speed. The trees on either side of the road are lush and green, with some leaves starting to change color, indicating that it might be autumn. The sky is clear and blue, suggesting a pleasant day. The overall style of the video is dynamic and energetic, capturing the motion of the car and the beauty of the surrounding environment."
+    ],
+    "video_ids": [
+        "xdlStiy4Tmc_17_0to162"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a black sports car, specifically a Chevrolet Camaro, driving down a road. The car is clearly visible and matches the description of a black sports car. Additional elements like pedestrians and cyclists in the background do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a black sports car driving which a sleek Camaro, which on a road. The car is the the and matches the description of a black sports car. The elements such the and trees are the background do not contradict the core description of}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..42f3d2a85a063a6659d50e691265fa529f5f0819
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e57578b8c8a12daa8f48035dc3fba0c1a1aaa154f36c4e2caccf56249dbd054d
+size 383264
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..5b5e3c8f89fc1cbc6ab16d5df7ced1404d36cc08
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a 3D animated brain with a realistic texture and color, set against a blue background. The brain is shown in three different positions, each time with a slightly different angle and perspective. The animation is smooth and fluid, giving the impression of the brain moving or rotating. The style of the video is educational and informative, likely used to illustrate the structure and function of the brain."
+    ],
+    "video_ids": [
+        "RQVQrT3-vPU_0_0to193"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: 3D animated brain? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video displays a 3D animated brain rotating against a blue background, matching the description of a 3D animated brain. There are no conflicting elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts two 3D animated representation with against a blue background. which the description of a 3D animated brain. The are no additional elements in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5d95dbcecf8aa0d5a5b14d0eb6f39e81079d57ec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0ea4956323be48ac6a70cb81854f6ee96e9e0e85daf6e19a4beaf22b95dc366e
+size 296841
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..312202a4b417944aaf4df79c6055204abfb41c64
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a vibrant underwater scene within an aquarium. The setting is lush with green aquatic plants that vary in size and density, creating a dense, natural-looking environment. Among these plants, several colorful fish swim gracefully. The fish include neon tetras, recognizable by their bright red and blue bodies, and other smaller, more subdued-colored fish that add to the diversity of the scene.   In the background, a large, dark rock formation provides a focal point and adds depth to the aquarium's landscape. The water is clear, allowing for a detailed view of the fish and plants. The lighting appears to be soft and diffused, enhancing the natural colors of the fish and the greenery. The fish move fluidly through the water, occasionally interacting with each other or the plants, contributing to the dynamic nature of the scene.  As the video progresses, there is a subtle shift in perspective, suggesting a gentle pan across the aquarium. This movement allows viewers to observe different areas of the tank, highlighting"
+    ],
+    "video_ids": [
+        "2df95938dbedd356f72b4c81f154a1b578a4d3a67206ebb980dd3457dca46a53"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Neon tetras (bright red and blue bodies) and smaller, more subdued-colored fish.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows neon tetras with bright red and blue bodies swimming among other fish and aquatic plants. There are also smaller, more subdued-colored fish visible, such as the goldfish, which fits the description. The presence of additional elements like plants and rocks does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows neon tetras with bright red and blue bodies, among other fish. plants plants. The are also smaller, more subdued-colored fish present in which as the whitefish in which match the description of The presence of additional elements like the and a does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5b3f4c69cbedd9ca509d6a6032305c72f984f974
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f39237aaa952c2594132887acca055c6505d16932bdbad5811e072f74092db77
+size 311544
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..acd4aa7b49e416c4fc56db36836f3daef9d55892
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two men are standing at a wooden table, engaged in a conversation. The man on the left is holding a tablet, possibly showing something to the other man. The man on the right is gesturing with his hands, indicating that he is actively participating in the discussion. The setting appears to be a casual, indoor environment, with a potted plant and a window in the background. The overall style of the video suggests a friendly, informal interaction between the two men, possibly discussing a project or sharing information on the tablet."
+    ],
+    "video_ids": [
+        "-4s2MeUgduo_25_0to112"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, a wooden table, a tablet, and a potted plant.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men standing at a wooden table. A tablet is visible on the table and is interacted with by one of the men. A potted plant is also visible in the background. All core elements mentioned in the condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two men standing around a wooden table. One man is being in the table, one beinged with by one of the men. Additionally potted plant is also visible in the background, The elements elements of in the description are present and accurately depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a4b3c20c66913b6f75f81a507cf8290fc775cd8b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d4c33ad2f7e70336e379657c9578dacd23a0268d5047b7606d50982fe15793a2
+size 150265
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..6ed68fb0f7e5fad711dcd5afee7f0921bcd1c19f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person installing a car seat in a vehicle. The car seat is a black and gray infant seat. The person is wearing a blue shirt and green shorts. The car interior is white. The car door is open. The car is parked. The person is standing outside the car. The car is a luxury vehicle. The car seat is being installed in the back seat of the car. The person is using a tool to install the car seat. The car seat is being secured with a seat belt. The car seat is being installed in the middle of the back seat. The person is standing behind the car seat. The car seat is being installed in the car. The person is standing in front of the car seat. The car seat is being installed in the car. The person is standing in front of the car seat. The car seat is being installed in the car. The person is standing in front of the car seat. The car seat is being installed in the car. The person is standing in front of the car seat. The car seat is being installed in the car. The person is standing in front of the car seat. The car seat is being installed in the car. The person is standing in front of the car seat. The car seat is being installed in the car. The person is standing in front of the car seat. The car seat is being installed in the car. The person is standing in front of the car seat. The car seat is"
+    ],
+    "video_ids": [
+        "D_EghGVvjP8_15_0to146"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a black and gray infant car seat, a tool, and the car.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person installing a black and gray infant car seat in a car, which matches the core description. The person is clearly visible, the car seat is black and gray, and the car interior is shown. No contradictory elements are present that would invalidate the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person interacting a black and gray infant car seat into the car. which ful the ' description. The person is using visible, and car seat is black and gray, and the car is is visible. There additional elements are present.\"\n would suggest the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fdeef04d312afb185f490fa8b0a7d028197b7246
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f76ad4a78664e51d4a658f72cbd1cc74e0ed0979a06f183b957033e712ee29c9
+size 180793
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..64e0b4bb43dfb68a492c47020b1675d56e06577b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment on a football field. A group of football players, dressed in white and red uniforms, are engaged in a play. The players are in various positions, some are running, while others are in a defensive stance. The action is intense, with players colliding and pushing each other. The field is marked with white lines, indicating the boundaries of the game. In the background, a brick building can be seen, adding to the urban setting of the game. The video is shot from a high angle, providing a comprehensive view of the action on the field. The style of the video is realistic, capturing the intensity and excitement of the sport."
+    ],
+    "video_ids": [
+        "Kw32yRi0zFc_10_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A group of football players wearing white and red uniforms.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group of football players on the field, and they are wearing uniforms that are predominantly white with red accents, which matches the description. The players are engaged in a play, and their uniforms are clearly visible, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a group of football players wearing a field, and they are wearing white that are predominantly white with red accents, which matches the description of The players are engaged in a typical, which their attire are clearly visible, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..407e1f9a2f3f13befc7af5073b3fefad73c3ce97
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e67724cf1f55e6a17fea1bcbfee4322d634a015d87cf8d0b86a6b49526cb8698
+size 179646
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..7db875b642a900c2a84921eecb938a679c5e960b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a scene in a garage where two cars are parked. The first car, a sleek black sports car, is positioned on the left side of the frame. Its shiny exterior reflects the ambient light in the garage. The second car, a dark blue sedan, is parked on the right side of the frame. Its polished surface also reflects the light in the garage. The garage itself is spacious and well-lit, with a concrete floor and a white wall in the background. The cars are stationary, suggesting a quiet moment in the garage. The video does not contain any text or additional objects. The focus is solely on the two cars and their surroundings."
+    ],
+    "video_ids": [
+        "Kyus8d9TfMs_9_92to243"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two cars - a sleek black sports car and a dark blue sedan.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two cars: a dark blue sedan in the foreground (with its taillight visible) and a sleek black sports car (Aston Martin) in the background. Both cars are clearly identifiable and match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two cars, a sleek blue sedan and the background andwhich a rearillights visible) and a sleek black sports car inwithcura Martin) in the background. The cars match parked visible as match the description of.}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a651e4c71634abd144727b42b99c6c07b5b16bb8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1173-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:135391693c340d1c7b61d7cce011b129cbdbe379915829410504cfd9e810843d
+size 67949
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..fc98f70196e48773497fef535b3ca031092d2275
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a row of canned beverages, likely beer, being dispensed from a machine. The cans are shiny and metallic, with a blue and silver color scheme. The cans are lined up in a row, and the machine is in the process of dispensing them. The cans are being dispensed one by one, and the machine is in the process of opening the cans. The cans are being dispensed into a metal container, which is likely a part of the machine. The machine is likely located in a bar or a similar establishment. The style of the video is a close-up shot, focusing on the cans and the dispensing process. The video does not show any people or other objects, and the focus is solely on the cans and the machine. The video does not show any text or other graphics. The video is likely intended to showcase the dispensing process of the cans, and the focus is on the cans and the machine."
+    ],
+    "video_ids": [
+        "CRcxRMnUbp8_11_0to140"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Canned beverages (beer), machine, metal container? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a machine processing canned beverages, which appear to be beer cans, and includes metal containers as part of the machinery. All elements align with the described conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a machine that cans beverages, which align to be beer cans based as they a containers as part of the machine. The elements in with the ' conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5d1c7c3489a698a9ef2b8df16fbda47abaf730e2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:98d37472b94bb6240dfac90f7c457f4f363c01861fb334ec27ddc5a1310535c1
+size 231151
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..2b3034d6b392960e94760f9c4f88d7d87726aff3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D rendering of a kitchen interior, showcasing a variety of objects and furniture. The kitchen features a wooden floor and a purple door, with a large window allowing natural light to enter. A wooden counter with a sink is visible, and a coffee maker is placed on it. Above the counter, a shelf holds various items, including a vase and a bowl. A purple pendant light hangs from the ceiling, casting a warm glow over the counter. A rug is placed on the floor, adding a touch of color to the space. The overall style of the video is realistic, with attention to detail in the textures and lighting. The camera angle is from above, providing a comprehensive view of the kitchen layout and design."
+    ],
+    "video_ids": [
+        "0GeKa4vQpHQ_32_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Wooden counter with a sink, coffee maker, shelf with a vase and a bowl, and a purple pendant light.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a wooden counter with a sink, a coffee maker, a shelf with a vase and a bowl, and a purple pendant light. These elements are clearly visible and match the description provided, even though the video also includes other decorative items such as hanging lamps and stairs, which do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a kitchen counter with a sink, a coffee maker, a shelf with a vase and a bowl, and a purple pendant light. These elements match all visible and match the description provided. indicating though the video is includes a elements elements like as a plants and a, which are not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d8f0aef93e7a69c86078052a4ae6281d88574f8f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:16e9fc46d0f171f96adf62bac6c5580d700612cbf045832542ced12a7887b731
+size 82932
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..d5bfbc83bac0a886e07d727cac335a212c13d857
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, three astronauts are seen in a space station. The first astronaut, dressed in a blue and yellow suit, is standing with his hands behind his back, looking at the camera. The second astronaut, also in a blue and yellow suit, is standing with his hands on his hips, looking at the camera. The third astronaut, wearing a beige suit with a patch on the shoulder, is standing with his hands on his hips, looking at the camera. The space station has a green wall in the background."
+    ],
+    "video_ids": [
+        "TY0wIV2y7Hs_3_51to204"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three astronauts in different colored suits.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three men in what appear to be astronaut uniforms, each wearing a different colored suit: one in beige, one in blue, and one in maroon. These suits are distinct in color, fulfilling the condition of 'Three astronauts in different colored suits.'\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows two individuals dressed astronaut appear to be astronaut suits, each wearing a different colored suit: blue in blue, one in blue and and one in aoon. The suits are distinct in color and fulfilling the ' of havingdifferent astronauts in different colored suits.' The}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..39c059b96bb11a48aae5cdbd8ff5d7127d5b1ff1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40511a993b1617ea81ffb0fa4b7ab09e72652da055e0319db6bf789fcf421d9d
+size 160090
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..db84c4ea0bf01bf280d76516efaae9fcea7230bd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a classic car, showcasing its vintage design and features. The car's dashboard is adorned with a variety of gauges and dials, each serving a specific purpose in monitoring the vehicle's performance. The steering wheel, a prominent feature in the center of the dashboard, is made of wood and metal, adding a touch of elegance and sophistication to the overall design. The car's interior is bathed in sunlight, highlighting the details of the dashboard and steering wheel. The video is a testament to the craftsmanship and attention to detail that went into designing this classic car."
+    ],
+    "video_ids": [
+        "x7skkkY-xq4_50_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dashboard gauges and dials, steering wheel? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the dashboard gauges and dials, along with the steering wheel, which are the primary objects mentioned in the condition. The camera focuses on these elements, and they are prominently displayed throughout the video, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a dashboard withuges and dials, as with the steering wheel, which are the core elements mentioned in the description. The image angle on these elements, providing there are prominently displayed, the video. fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d2cb92a0c3e68d4602830e592cab99337c61ed89
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b3f0ea871faba94e352861257e106cd522db5ba9bb0eed8221926778aa81e85c
+size 129228
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..35f2b9c3c0568d20e5297b0724ec9427f04a30d9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a gray Audi car in a garage. The car is parked and stationary. The car has a license plate that reads \"LP ML 87\". The car has a sleek design with a shiny exterior. The garage has a white floor and a white wall. There is a sign in the background that reads \"Profis\". The car is the main focus of the video, and it is shown from a rear view. The video is a still image, and there is no movement or action. The style of the video is straightforward and simple, focusing on the car and its details."
+    ],
+    "video_ids": [
+        "iXCcS2_Kfp8_5_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A gray Audi car with a license plate reading 'LP ML 87'.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a gray Audi car from the rear, and the license plate is visible and reads 'LP ML 87'. The car's model and color match the description, and no elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a gray car car with the rear angle which the license plate is visible and reads 'LP ML 87'. The car's design and color match the description, and there additional contradict the information description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fcb4d20dda151aebbdebe83963d67a78152f896c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7ea83da31437f07850501e1ace6dc75d3f99007bae3d350da003008ab0a40f0f
+size 167086
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..f5d877114275f4fb250c79f2d8412bd4ac5c3a8d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen preparing a dessert, specifically a cake doughnut, on a wooden cutting board. The doughnut is placed on a white plate, and the person is using a spoon to drizzle white icing on top of it. The icing is being poured from a glass bowl. The setting appears to be a kitchen, and there is a fork on the side, suggesting that the dessert is ready to be served and enjoyed. The overall style of the video is simple and straightforward, focusing on the preparation of the dessert."
+    ],
+    "video_ids": [
+        "9u_YgDglbYs_52_0to113"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a cake doughnut, a white plate, a spoon, a glass bowl containing white icing, and a fork.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person drizzling white icing from a glass bowl onto a cake that resembles a ring-shaped doughnut, placed on a white plate. A spoon is used for drizzling, and a fork is visible on a small plate to the side. All specified objects are present and correctly depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person pouringizzling white icing from a spoon bowl onto a cake dough is a dough don doughnut. placed on a white plate. The spoon is used to theizzling the and a fork is visible on the wooden wooden in the side. The the objects are present and correctly described in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1758310bdfaff800cb1361fb06512899e51a81d7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:140f607c84c54055aae67d821ddb7ee7ab91680014fad4ebea5a65cf50e52b68
+size 133371
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..562a51756592eba6e1bd2c155c5ad1dd70a8e099
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a triumphant moment of a football player celebrating a victory. The player, wearing a gray cap and a white towel, is holding a trophy high above his head. He is surrounded by a shower of blue and green confetti, adding to the festive atmosphere. The player's smile is wide and genuine, reflecting the joy of the moment. The background is blurred, but it appears to be a stadium filled with spectators, suggesting that this is a significant event. The style of the video is dynamic and energetic, capturing the excitement and pride of the player and the crowd."
+    ],
+    "video_ids": [
+        "tyVrOg38EJ8_34_0to123"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A football player, a gray cap, a white towel, a trophy, blue and green confetti? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a football player holding a trophy, wearing a gray cap with 'CHAMPIONS' written on it, draped in a white towel, and surrounded by falling blue and green confetti. All core elements from the description are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a football player wearing a trophy, wearing a gray cap, aDICIONS' written on it, and with a white towel, and surrounded by blue blue and green confetti. The the elements of the description are present, accurately represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f6937f54a8935240a7a26397f906c6f5f15d36c7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6cf395f86613a6c57ca3e32a409f511e76a3c92d69db5cbc6c130346e41bcb7
+size 316655
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..5d21eed535fbb36f06477893fa754d19e0544357
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a vibrant orange race car in motion on a track. The car is adorned with blue and white stripes, and the number 75 is prominently displayed on its side. The car is equipped with a black racing seat and a roll bar for safety. The track itself is surrounded by lush green grass, providing a stark contrast to the car's bright orange color. The car's speed and the surrounding environment create a dynamic and exciting scene."
+    ],
+    "video_ids": [
+        "E1WKmm_6Cj4_4_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A vibrant orange race car with blue and white stripes, a black racing seat, and a roll bar.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a vibrant orange race car with blue and white stripes, which matches the description. The interior is visible through the window, showing a black racing seat and a roll bar (the safety cage structure). Although the car is in motion and there are other elements in the background, these do not contradict the core description of the car's appearance and interior features.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a vibrant orange race car with blue and white stripes, which matches the description. The car of not, the car, and a black racing seat and a roll bar,though roll bar around). The the video is moving motion, the are some elements like the background, the do not contradict the core description of the car's appearance and features.\"\n.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0fda821bd4caed28ccee9f304fe00b6e67a83a9f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1174-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3f9d1d2c5cabea1b8e4315f6a01a432b387690ca4efa5318ac41d8da02f29d34
+size 156955
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..462b0fbaa3ab75629372382625713fc01265bc60
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a vibrant and dynamic collage of superheroes from the Marvel Cinematic Universe. The style is a blend of action and fantasy, with each superhero showcasing their unique powers and costumes. The collage is set against a backdrop of a colorful explosion, adding to the excitement and energy of the scene. The superheroes are arranged in a way that creates a sense of movement and dynamism, with some characters appearing to be in motion while others are in a more static pose. The overall effect is a visually stunning representation of the Marvel superheroes, capturing their individual strengths and the camaraderie that unites them."
+    ],
+    "video_ids": [
+        "6SA2oQlLStA_22_0to177"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Superheroes from the Marvel Cinematic Universe? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a collage of well-known superheroes from the Marvel Cinematic Universe, including Thanos, Gamora, Black Widow, Scarlet Witch, Vision, Doctor Strange, and others, all depicted in their iconic costumes and poses, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The image features a group of characters-known superheroes from the Marvel Cinematic Universe, including charactersos, Captainora, Iron Widow, Captain Witch, and, and Strange, and others. all depicted in their iconic costumes. poses. which the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8fb7c142cd6f55d6a66c4fff9a7ecea65a14d4ca
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e3346f57f9c349b4cb779c5d7c08718d9792470f932361057550f6d0cc1131a9
+size 106827
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..5b52bb4b3249abba81f7df87874c3a0bf0159765
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is an aerial view of a desert landscape, showcasing a large, well-maintained property with a variety of structures and greenery. The property features a central building with a red roof, surrounded by a series of smaller buildings and green spaces. A paved driveway leads to the central building, where a white car is parked. The property is bordered by a fence, and there are several trees and bushes scattered throughout the area, providing a stark contrast to the otherwise arid surroundings. The sky is clear and blue, suggesting a sunny day. The style of the video is realistic, with a focus on the property and its surroundings. The video does not contain any people or animals."
+    ],
+    "video_ids": [
+        "tKOUnPX2Z1w_8_0to157"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Central building with a red roof, smaller buildings, green spaces, paved driveway, white car, fence, trees, and bushes.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a central building with a red roof, surrounded by smaller buildings, green spaces, a paved driveway, a white car, a fence, and trees and bushes. All these elements are clearly visible and align with the described condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a suburban building with a red roof, smaller by smaller buildings, green spaces, paved paved driveway, a white car, a fence, and trees and bushes. The the elements are present visible and match with the description scene.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6a8be7ccc0d642e8b5759014421d962e60d6ef06
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:22888e9c02de653032d6708de691ab8b29094635dde375f4dd27196e2d5b0845
+size 255002
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..11eeb36ce81448a45a659aaf480820e7cb432ce4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a car's air vent, which is located on the side of the car door. The vent is circular in shape and has a black and yellow color scheme. The yellow ring around the vent adds a pop of color to the otherwise black vent. The vent is made of plastic and has a shiny finish. The car door appears to be made of leather, adding a touch of luxury to the scene. The vent is the main focus of the video, with no other objects or actions taking place. The style of the video is simple and straightforward, focusing solely on the vent and its details."
+    ],
+    "video_ids": [
+        "WjKneIxybNc_37_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Car air vent? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a car air vent with a distinctive neon green ring, embedded in the car's interior panel. The object is accurately depicted as a car air vent, matching the specified condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a close air vent, a circular design yellow and and which in a dashboard's dashboard.. The design in identifiable depicted as a car air vent, fulfilling the description '.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9293e1ca42adbca387edaeb684f122d6cf162070
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f8426a30c620ee72ff90b96d525cda651c48bb5c2b7045e04730f3a8d4d90e1b
+size 197349
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..00c8798cdaa0c906dd1537517a11c95c9ba11cb6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young girl with blonde hair is seen in a whimsical outdoor setting. She is dressed in a purple outfit and holds a butterfly net in her hand. The girl is standing on a dirt path, surrounded by tall sunflowers that tower over her. The sun is shining brightly, casting a warm glow on the scene. The girl appears to be in a joyful mood, as she is smiling and looking towards the camera. The overall style of the video is cartoonish and colorful, with a focus on the girl and her interaction with nature."
+    ],
+    "video_ids": [
+        "-jbiWXjorcc_195_18to150"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young girl with blonde hair, dressed in a purple outfit, holding a butterfly net; tall sunflowers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young girl with blonde hair, wearing a purple outfit and holding a butterfly net, which matches the core description. Tall sunflowers are also visible in the background, fulfilling that part of the condition. The presence of a caterpillar and trees does not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a young girl with blonde hair, dressed a purple outfit, holding a butterfly net. which matches the ' description. Additionally sunflowers are also present in the background, fulfilling the part of the description. The presence of the brightpillar on the does not contradict the description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6675b7bc2174aed1e9a582318d52f9023da8982f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40484a1fd087804dfbde3b7fd48739825ae59bada0ee77218aaef9c4a6793b01
+size 94363
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..d3718a0d3db0d5502899c095d67217f4f90012f7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a woman standing on a rocky hilltop, overlooking a breathtaking landscape. She is dressed in a white shirt and black pants, and she holds a camera, seemingly capturing the stunning view. The hilltop is covered in lush green grass, and the rocky terrain adds a rugged charm to the scene. In the distance, the ocean stretches out, its blue waters contrasting beautifully with the green of the hilltop. The sky above is a clear blue, dotted with fluffy white clouds. The woman's position on the hilltop gives her a commanding view of the landscape, and her presence adds a human element to the natural beauty of the scene. The video is a beautiful blend of nature and human interest, capturing a moment of tranquility and awe in the face of the natural world."
+    ],
+    "video_ids": [
+        "9Va_JPlGvhY_50_95to249"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a rocky hilltop, lush green grass, the ocean, and fluffy white clouds.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a woman standing on a rocky hilltop, surrounded by lush green grass, with the ocean visible in the distance and fluffy white clouds in the sky. All specified elements are present and accurately depicted without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a woman standing on a rocky hilltop, with by lush green grass. with the ocean in in the background. fluffy white clouds in the sky. The the elements are present and contribute depicted, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..21afe0504cfacaa50ad4a49f48ad97c70e5bdaac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4aac69d5da2a6294d58fb0d33c9dfd8270e9812e994610f75d6c143460f2cd91
+size 254235
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..df463558920c81cd80865624e2625421c9ed4b1c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a 3D animated character, a firefighter, standing next to a red fire truck. The firefighter is wearing a yellow helmet with a red star on it, a blue uniform with silver buttons, and a green face mask. The fire truck has a hose attached to it, and the firefighter is holding the hose. The background shows a brick building and a house with a chimney. The firefighter is looking at the camera with a serious expression. The style of the video is cartoonish and colorful, with a focus on the firefighter and the fire truck."
+    ],
+    "video_ids": [
+        "KkpWfTPB_z0_279_482to662"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ['firefighter', 'red fire truck', 'brick building', 'house with a chimney']? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a firefighter in uniform, a red fire truck partially visible behind him, a brick building (fire station) to the right, and a house with a chimney in the background on the left. All specified objects are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts two scene in a, a red fire truck, visible in the, and brick building,which station), in the left, and a house with a chimney in the background. the left. The elements objects are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8eb93f1e9c06dffac76c9ca224409821340db141
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4c15b997463a9eb7e9da001fbfd39535aabfdfbbffd75d4567458bb0405d20e5
+size 119366
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..ae09b9d9ac71eb3eb171b8d682fef93c1b88ad48
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a bouquet of flowers, specifically roses, in a vase. The roses are in full bloom, with their petals appearing soft and delicate. The bouquet is set against a wooden table, which provides a natural and rustic backdrop. The lighting in the video is soft and warm, highlighting the beauty of the flowers and creating a serene atmosphere. The focus of the video is on the roses, with the background being blurred, drawing the viewer's attention to the bouquet. The style of the video is simple and elegant, with a focus on the beauty of nature."
+    ],
+    "video_ids": [
+        "1TyOfqcULVc_45_0to168"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bouquet of roses in full bloom? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bouquet of artificial roses in full bloom, with visible petals and a natural arrangement. While there are additional elements like small white berries and green leaves, they do not contradict the core description of a bouquet of roses in full bloom. The roses are the central focus and appear as described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a close of roses roses in full bloom, which a petals and green natural arrangement. The the are no elements like leaves green flowers and green leaves, they do not contradict the core description of a bouquet of roses. full bloom. The focus are the main focus, fulfill to if.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1b539834e73b362b0efb64c942db7fb4ee701a27
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d28f9ffac4b08d2d23babbdd5f69ddd4615c3f69e4f661575ab285d69bee7b18
+size 111126
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..737ccd88b08af1768ca6e91516d191372698d306
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene beach scene, with a variety of seating options available for visitors. The beach is adorned with several straw umbrellas, providing shade for those seeking respite from the sun. Underneath these umbrellas, there are numerous blue and white lounge chairs, inviting beachgoers to relax and enjoy the view. The beach is also equipped with a few tables, perfect for enjoying a meal or a drink. The overall atmosphere is one of tranquility and leisure, with the inviting furniture and the calming sound of the waves in the background."
+    ],
+    "video_ids": [
+        "K3SFFuIXosc_69_0to129"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Straw umbrellas, blue and white lounge chairs, tables.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows multiple straw umbrellas providing shade over lounge areas. There are also numerous blue and white lounge chairs and tables arranged on the beach, matching the described objects. The scene is consistent with a beach setting, and no elements contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts straw straw umbrellas and shade, blue chairs. The are blue blue blue and white lounge chairs visible tables visible on a sandy, which the description elements. The scene is set with the typical setting, and the additional contradict the given.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fa5da683fd9780b112e7bdddb11e1c2bc6cf7e2a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1175-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae4689525e2c650dea8c2b5bc541532d12616b21b488e3e35d4a63622c154b99
+size 151513
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..84f3809e36e81877fd9af9834d0325de1612efc2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a cooking tutorial, featuring a close-up view of a table with various ingredients and tools. The table is covered with a pink cloth, and there are three baskets on it. The first basket contains a dozen eggs, the second basket contains a dozen sushi rolls, and the third basket contains a small bowl of mayonnaise. There is also a blue lid on the table. The style of the video is simple and straightforward, focusing on the ingredients and tools used in the cooking process. The video is likely to be informative and educational, providing viewers with a step-by-step guide to making sushi rolls."
+    ],
+    "video_ids": [
+        "cZBZxoWA64g_24_0to158"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three baskets (first basket: dozen eggs, second basket: dozen sushi rolls, third basket: small bowl of mayonnaise), a blue lid? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three baskets: one with eggs, one with sushi rolls, and one with a small bowl (likely mayonnaise). A blue lid is also visible on the left side of the frame. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three baskets: one with brown, another with what rolls, and one with a small bowl ofwhich containingonnaise). There blue lid is also present. the table side of the frame. The elements elements of in present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e2c8f815d196ad976eb4aae2975d9bd33efbc490
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:537ffbaa94a8d0f983db6f68e6421dda04534022835b86fadc2756401a4e1cea
+size 126137
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..4cddd80518e3a30c63b2a1bcc08d22b85e7db7a7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the breathtaking view of a rocky coastline under a clear blue sky. The camera pans across the scene, showcasing the towering cliffs that rise majestically from the deep blue water. The cliffs, with their rugged texture and jagged edges, are a testament to the power of nature. The water, a deep shade of blue, is calm and serene, reflecting the clear sky above. The video is a beautiful representation of the natural beauty of the coastline, capturing the essence of the location in a single frame."
+    ],
+    "video_ids": [
+        "TGkmMusQL9g_24_0to143"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Cliffs, water, sky? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features cliffs rising from the water, a vast expanse of water in the foreground, and a clear blue sky with scattered clouds. These elements fully satisfy the 'Object(s)' condition as described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features cliffs, sharply the water, with clear expanse of blue, the middle, and a clear blue sky. no clouds. These elements collectively fulfill the 'Cl(s)' condition as described.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d9366a76007c73a67e4494b6b29fdde6a333b228
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d74d160990f63b06008482cce65e89766d0c8e8f2de6ec8f3853f420cdc4967b
+size 118396
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..094ad242085c3ee1b12313dafce6ae64e6094702
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman in a pink shirt standing in front of a toaster oven. She is gesturing with her hands, possibly explaining how to use the appliance. The toaster oven is placed on a surface, and the woman appears to be in a kitchen or a similar setting. The style of the video is likely instructional or promotional, aiming to showcase the toaster oven's features or usage. The woman's attire and the setting suggest a casual and approachable tone for the video."
+    ],
+    "video_ids": [
+        "noEfrWETyv4_1_27to192"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a pink shirt and a toaster oven.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a pink shirt standing next to a toaster oven, which matches the core description. The background is plain white, and there are no conflicting elements that contradict the specified objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a pink shirt standing in to a toaster oven. which contains the description description. The presence and a and, and there are no additional elements. contradict the description objects.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2e011eaf552d25abdca4c67d878120092ac1ac0a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4bb92fc4aa1a1c2420fecbd34df5c1e7e5bae99d6aa6a53fee88f9f0d4544729
+size 139638
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..f5bbf282c865146c6d146f4b7a3d9344caccfb03
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a car engine, focusing on the engine's components and the exhaust system. The engine is silver and blue, with various parts such as the air intake, throttle body, and exhaust pipe visible. The engine is mounted on a black engine mount, and the exhaust pipe is connected to the engine with a clamp. The engine is in a stationary position, and there is no visible movement or action in the video. The style of the video is a straightforward, unembellished documentation of the engine and its components, with no additional context or narrative provided."
+    ],
+    "video_ids": [
+        "AfgXoCr-gzg_16_53to175"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A silver and blue car engine with visible components such as the air intake, throttle body, and exhaust pipe. The engine is mounted on a black engine mount, and the exhaust pipe is connected to the engine with a clamp.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a car engine with visible components including a silver air intake system, a throttle body, and an exhaust pipe connected with a clamp. The engine is mounted on a black engine mount, and the overall color scheme includes silver and blue elements. These details align with the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a car engine that silver components such the silver and intake,, a throttle body, and an exhaust pipe. with a clamp. The engine is mounted on a black engine mount, which the overall color scheme is silver and blue,, The details align with the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..445d89cb95b6cfc5f5a43f06bcc8be911994dfb5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf3d10460a43472961ff2446abd5cc30db8b8a642a9efe700aaf560b0b0f0b9a
+size 69236
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..3df00079037882aeaa0e3c622b7b0083b3040207
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a quaint, rustic wooden cabin nestled in a lush, verdant forest. The cabin, with its thatched roof and stone accents, exudes a sense of tranquility and simplicity. The viewer's perspective is from the ground, looking up at the cabin, which is surrounded by towering trees and dense foliage. The sunlight filters through the canopy, casting dappled shadows on the cabin's exterior. The overall style of the video is serene and naturalistic, capturing the essence of a peaceful retreat in the heart of the woods."
+    ],
+    "video_ids": [
+        "m9FmEygX35E_2_21to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A rustic wooden cabin, towering trees, dense foliage? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by prominently featuring a rustic wooden cabin with a round door and stone accents, surrounded by towering trees and dense foliage. The cabin is nestled in a forest setting, with visible tree trunks and lush green leaves, matching the description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition by depicting featuring a rustic wooden cabin, a triangular window and a foundation, surrounded by towering trees and dense foliage. The setting is nestled within a forested, which the green trunks and green greenery, which the description provided any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..41a18b2873ef8541784af89252839d2014cfe7a3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7c25961c4ba6de8301573fd8a67d8eca98c49bc7acd1e24a4a52e69afdb3d6b2
+size 211393
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..59341096c3a01f3d39e5bfca5e62fa4ce03d3683
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene walk through a lush tropical garden. The viewer follows a winding path lined with towering palm trees and a variety of vibrant green plants. The path is bordered by a low stone wall, adding a touch of rustic charm to the scene. The garden is bathed in soft, natural light, highlighting the rich colors and textures of the foliage. The overall style of the video is tranquil and inviting, offering a glimpse into the beauty of nature."
+    ],
+    "video_ids": [
+        "eUcC48zRUxY_4_0to200"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Towering palm trees, vibrant green plants, a low stone wall? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features towering palm trees and vibrant green plants, which are prominently displayed throughout the lush, tropical setting. A low stone wall is also visible along the pathway, separating the walkway from the surrounding foliage. These elements align with the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a palm trees, vibrant green plants, which are clearly displayed. the scene garden tropical setting. The low stone wall is also visible, the pathway, separating the plantsway from the dense green. The elements collectively well the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..37499cdb19575f799f851d82ef271efa7241cddc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:427cbddfcd26d5b71477f6fa63bc922d97fbda49cbc228c716b43f0c63c3b47d
+size 347371
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..e2dcd3e71cf69c92d2b698d6f03adba002455b1c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a pumpkin pie in a tin foil pie pan, placed on a purple napkin. The pie has a golden-brown crust and a creamy yellow filling. A green leaf garnish is placed on top of the pie. In the background, there is a NutriBullet blender with a silver base and a clear pitcher. The style of the video is a simple, straightforward food presentation, focusing on the pie and the blender. The lighting is bright, highlighting the textures and colors of the pie and the blender. The video does not contain any text or additional elements."
+    ],
+    "video_ids": [
+        "QxeVuSoFHFw_11_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Pumpkin pie in a tin foil pan, green leaf garnish, NutriBullet blender? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a pumpkin pie in a foil pan, garnished with a green leaf, and a NutriBullet blender is visible in the background. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a pumpkin pie in a tin pan, garnished with a green leaf, and a NutriBullet blender in visible in the background. The elements elements of in present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..75ee6edbdfe31de4adeb0355a0726f8ce2dea531
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9021ce2e3276f442ae018fdcc3ac32cdccc84815ec54053ca57820fb315de60d
+size 75941
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..e6a33fe8e9a93ad716e439e4eb86acf4cb57755a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a stylized, animated depiction of the superhero Captain America. He is shown in three frames, each capturing a different pose and action. In the first frame, he is seen holding a shield with the American flag design, standing confidently with his arms crossed. In the second frame, he is shown in a dynamic pose, running forward with his shield held in front of him. In the third frame, he is seen in a fighting stance, ready to strike with his shield. The background of each frame is a vibrant, red, white, and blue American flag, emphasizing the patriotic theme of the character. The style of the video is reminiscent of comic book art, with bold lines and vivid colors."
+    ],
+    "video_ids": [
+        "NJB_n7XrPNc_18_18to198"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Captain America, shield? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features Captain America in his iconic costume, holding his signature shield. The visual elements are consistent with the description, and no conflicting elements are present that would contradict the core description of Captain America and his shield.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a America, various iconic costume, holding a shield shield. The presence elements, consistent with the description, showing the additional elements are present. would contradict the ' description of the America and his shield.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ea149eb6054824f863fe80d5b5a5f3a5c7b8cfe6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1176-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8bc8a1c3463dc77f7ec1911de36c2c1d89720da1e7586ddc95b3fedf97085d07
+size 222748
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..6c1a14384bb0c69c18639757bef146d6880dae49
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a collection of cacti and succulents arranged in a wooden box on a table. The cacti and succulents are of various sizes and colors, including shades of green, blue, and purple. The arrangement is carefully crafted, with the plants placed in a way that showcases their unique shapes and textures. The wooden box provides a rustic and natural backdrop for the plants, enhancing their beauty. The table on which the box sits is made of wicker, adding to the overall natural and organic feel of the scene. The video captures the tranquility and simplicity of the arrangement, highlighting the beauty of these desert plants in a domestic setting."
+    ],
+    "video_ids": [
+        "cFtYeTng4OU_3_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Cacti and succulents? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a variety of cacti and succulents arranged in a decorative wooden tray. The plants display diverse shapes, colors, and textures typical of succulents, including rosette-forming species and small, spiky cacti. There are no elements that contradict the description of cacti and succulents as the main subject.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features succ variety of succacti and succulents, in a wooden wooden box. The plants are a shapes and sizes, and textures, of succulents, which shadessettes formationsing leaves and others, spiky cacti. The are no elements in contradict the description of cacti and succulents being the main focus of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0bf18cdd689e5fd846a4ff876159777e7c62afc8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dc346d0b547e902b9663c3cbb28e1fd3684b8689c6d230b71efd02083bf3f6f3
+size 190097
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..0830c848c5d1221c91a1a3c33f9bacbaa24c41fd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a plate of food, specifically a stack of pancakes with a drizzle of syrup and a sprinkle of lemon zest. The pancakes appear to be golden brown and fluffy, with the syrup adding a glossy sheen to the dish. The lemon zest adds a pop of color and suggests a tangy flavor to the dish. The plate is white, which contrasts with the darker colors of the pancakes and syrup, making the food stand out. The style of the video is simple and straightforward, focusing on the food without any additional context or background. The camera angle is slightly above the plate, providing a clear view of the food and its details. The lighting is bright and even, highlighting the textures and colors of the pancakes and syrup. The video does not contain any text or additional elements, keeping the focus solely on the food."
+    ],
+    "video_ids": [
+        "9afC4l2FQI0_27_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A stack of pancakes, drizzled with syrup, and sprinkled with lemon zest.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a stack of pancakes on a white plate, drizzled with syrup and sprinkled with lemon zest, which matches the description. The visual elements are consistent with the requested object(s) without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a stack of pancakes that a plate plate, whichizzled with syrup, sprinkled with lemon zest. which matches the description provided The syrup elements in consistent with the given object(s). condition any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..717013694bf237c60c9b54b6baa75c85cbebf6da
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0e13ffa523aa3988f5deb9eead54299a4c0883cce2012ae25c720c84dc87e9d3
+size 69115
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..9f21a8b86b5deb2e1481d3006cc2f2cbc3d0081b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman standing in front of a body of water, possibly a lake or a river, with a cityscape in the background. She is wearing a white blouse and a dark cardigan, and she has blonde hair. The woman appears to be speaking or presenting, as suggested by her facial expression and the microphone visible in the image. The style of the video is a straightforward interview or presentation, with the focus on the woman and her surroundings. The setting is outdoors, and the weather appears to be clear and sunny. The image quality is good, with natural lighting and clear visibility of the subject and the background."
+    ],
+    "video_ids": [
+        "EoPKtVFpd00_5_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a white blouse, a dark cardigan, blonde hair, and a microphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a woman with blonde hair, wearing a white blouse and a dark cardigan, and a microphone is visible clipped to her cardigan. These elements match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a woman with blonde hair wearing wearing a white blouse and a dark cardigan, standing speaking microphone is visible in to her cardigan. The elements match the description provided. the questionObject(s)' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b3cb417576e782aa1cdd5b7fb511abd6b80069c8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a85b16d60bccff500911d445d44e0267880d5013e678ea20c6e8313cad36c272
+size 177413
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..ae7cfcf82ea1c2f4c817dae926e33a18c6eb6b3d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a brown dog on a boat, wearing an orange life jacket. The dog is standing on the boat, looking out over the water. The boat is moving through the water, with the waves gently rocking the boat. The dog's life jacket is bright orange, providing a stark contrast to the dog's brown fur. The water around the boat is a deep blue, and the sky above is a clear, bright blue. The boat is moving towards the horizon, where the sky meets the water. The dog appears to be enjoying the ride, looking out over the water with interest. The overall style of the video is a simple, straightforward capture of a moment in time, with no additional embellishments or effects. The focus is solely on the dog and its surroundings, providing a clear and unobstructed view of the scene."
+    ],
+    "video_ids": [
+        "AKjaBgvuTBA_9_0to142"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A brown dog, an orange life jacket, a boat? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a brown dog wearing an orange life jacket, standing on a boat. The dog is the central subject, and the boat and life jacket are visible and consistent with the description. Additional elements like the water and sky do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a brown dog wearing an orange life jacket, standing on a boat. The dog is positioned main focus, and the boat is the jacket are visible. match with the description. The elements like the ocean and the do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ee21b743e42796417bdc626dedf61834c3f0c09f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab320975501d09a8fe210335e3917974debba4a9d2f43d708edb715e6c9e0eb0
+size 252413
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..0b7c82dec443d2971abc6397ec7169ea0828df0f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a kitchen, holding two martini glasses filled with ice. She is dressed in a black top and has her hair styled in a braid. The kitchen is well-lit and features a counter with various items on it, including a potted plant and a vase. The woman appears to be in the process of preparing the drinks, as she holds the glasses up to the light, possibly checking the ice content. The overall style of the video is casual and homey, with a focus on the woman and her actions in the kitchen."
+    ],
+    "video_ids": [
+        "CFYTC6OfN5U_7_33to215"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, two martini glasses filled with ice, a potted plant, and a vase.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman holding two martini glasses filled with ice. A potted plant (with red berries) is visible on the left side of the frame, and a vase (or vase-like container) is also present holding the plant. All core objects described are present in the video without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a woman holding a martini glasses filled with ice. There potted plant islikely green flowers) is visible in the counter side of the frame, and a vase withwith similar-like object) is seen present in yellow plant. The elements elements are in present and the video, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8c6b90b1e3edf3487cbae112625364991ff94a96
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf3b8b29ef44688b0951b49b226f27a6463ca5ee738829e8898cabe999a115d5
+size 151646
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..d75bf5ed0439e4b0782feba3c6320997778a456d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated scene depicting a vibrant and colorful landscape. It features a body of water with floating lily pads and flowers, and a large rock formation with flowers growing on it. The rock formation is connected to the ground by green vines. In the background, there are green hills and trees, and a small village with buildings and windmills. The sky is clear and blue, and the overall atmosphere is serene and picturesque. The style of the animation is bright and cheerful, with a focus on nature and the beauty of the environment."
+    ],
+    "video_ids": [
+        "Dxm9-sCSAZo_38_0to133"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Body of water, floating lily pads and flowers, large rock formation, green vines, green hills, trees, small village, buildings, windmills? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a body of water with floating lily pads and flowers, large rock formations adorned with green vines and colorful flowers, green hills, trees, and a small village with buildings and windmills in the background. All elements described in the condition are present and consistent with the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a vibrant of water with floating lily pads and flowers, a rock formations, with green vines, flowers flowers, green hills, trees, small a small village with buildings and windmills in the background. The elements match in the condition are present and contribute with the scene shown}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ed96edf8ab53e6b2f0cc4cb7c6d4cbb89a63af3d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd5054b38e3fbadb7d1577a3ab6422b3c83fbffde48dac8f2ef7d145385d21da
+size 257685
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..2cab4d32b694ca06302fa9349e9e47bd230a1cab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a vibrant underwater scene featuring a school of yellow and black striped fish swimming around a coral reef. The fish are darting and darting around the coral, their bodies shimmering in the light filtering through the water. The coral, a mix of orange and red, provides a stark contrast to the blue of the water. The fish are seen swimming in different directions, some closer to the camera, others further away, creating a sense of depth and movement. The video is a beautiful representation of marine life, showcasing the diversity and beauty of the underwater world."
+    ],
+    "video_ids": [
+        "G56T3AMvbG8_166_0to150"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A school of yellow and black striped fish? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a school of fish that are predominantly yellow with black stripes, matching the description. The fish are swimming around a coral structure, and while there are other elements like coral and possibly other fish, the main focus is on the school of yellow and black striped fish, which is accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a school of yellow with are predominantly yellow with black stripes, which the description of The fish are swimming in a coral reef, which the there are other fish like the and additional other fish species the main focus is on the yellow of yellow and black striped fish, which ful the represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1f7692c49e9e9c3bf4f92ff70905311cec0bd1af
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b4cbe9a5108cc4c86e8ec18c4f41c36f3363872de71e965f284654036ddad31
+size 338055
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..4e01a13f25356e658d357765ddab0998337b0d33
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman dressed in a green costume with a matching green wig and bow. She is wearing a green dress and has green makeup on her face. The woman is standing in front of a white staircase with a green railing. The background is blurred, but it appears to be an outdoor setting with trees. The woman is looking to her left with a surprised expression on her face. The style of the video is a mix of costume and performance, with a focus on the woman's outfit and expression."
+    ],
+    "video_ids": [
+        "cp1T5rODcuc_25_0to175"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a green costume and wig.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a woman wearing a green wig with pigtails, a large green bow, and a shiny green costume, which matches the description of 'A woman in a green costume and wig.' Additional elements like makeup and background do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a woman with a green costume and aigtails, a green green bow, and a green green dress with which align the description of aA woman in a green costume and wig.' The elements like the and accessories details not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..387751efe4f1e054a73b22f6290cb1c13f5ed83e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1177-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb3da9c7cfeac9e398bce3beb46570e33a09dd4eb5db09acc0b5e3e7c15ef15c
+size 202724
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..11f3d455740fa3f939455f7134dfde6564b55728
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men in suits, each raising their hands in a gesture of agreement or approval. The first man, on the left, is slightly ahead of the second man, who is on the right. Both men are wearing glasses and blue ties. The background is blurred, but it appears to be an indoor setting with a wooden paneled wall. The style of the video is a straightforward, unembellished news report, likely from a financial or business news source, as indicated by the \"Bloomberg\" logo in the bottom left corner. The focus is on the men and their actions, with no additional context or embellishments provided."
+    ],
+    "video_ids": [
+        "LLBRVZk9Ynk_4_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men in suits, both wearing glasses and blue ties.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men in suits, both wearing glasses and blue ties, as described. The scene appears to be a formal setting, possibly a hearing or deposition, where both individuals are raising their hands, consistent with the description. No elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows two men in suits, both wearing glasses and blue ties, which described. They men includes to be a formal setting, possibly a meeting or a, where individuals individuals are raising their hands, which with the ' of The additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c36eae1344735a9a6a9c0cac6da372d4a72f5536
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2a7a52aab3ab85babe3ae50df74864efab45d0301fb172493eaeaae3940ff337
+size 181662
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..c3c974a40b1ff12d97f82f60a6ce6eed8064885f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red and black motorcycle parked on a gravel road. The motorcycle is positioned in the center of the frame, with the front facing the camera. The background features a wooden fence and trees, suggesting a rural or semi-rural setting. The lighting in the video is natural, indicating that it was taken during the day. The style of the video is straightforward, with no additional elements or effects. The focus is solely on the motorcycle, highlighting its design and color."
+    ],
+    "video_ids": [
+        "4zl7jnt1He0_2_0to177"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red and black motorcycle? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red and black motorcycle, specifically an Indian model, parked on a gravel surface. The motorcycle's color scheme and design match the description, and no elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a red and black motorcycle, which a April F, which on a paved path. The motorcycle is design and and design are the description provided and there additional in the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..33c3aa7419b8d107d0603b9bdcb69f925cf1e4e6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8fdba67b623b3cf0228675312e4113c39e7b109b54499270e5aafb7c4ea0f2bb
+size 152798
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..91ddc576cbf0d8da51285742c2eca444e798d82e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a bicycle wheel in motion, showcasing its design and details. The bicycle is equipped with a black frame and a tan wheel, which stands out against the urban backdrop. The wheel is adorned with a distinctive logo, adding a touch of personality to the bicycle. The bicycle is parked on a street, with cars and buildings visible in the background, suggesting an urban setting. The video is shot from a low angle, emphasizing the bicycle's wheel and making it the focal point of the scene. The motion of the wheel suggests that the bicycle is in motion, adding a dynamic element to the otherwise static urban landscape. The video is a blend of stillness and movement, with the bicycle's wheel serving as a symbol of freedom and adventure in the heart of the city."
+    ],
+    "video_ids": [
+        "UJQ91Mg9Zf0_25_29to182"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bicycle wheel? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a bicycle wheel, showing its tire, rim, spokes, and hub in detail. The wheel is the central focus of the video, and while there are background elements like houses and vehicles, they do not contradict the core description of the bicycle wheel. The video successfully fulfills the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a bicycle wheel, which a details, rim, and, and part. detail. The wheel is the central focus, the video, and there there are other elements like a and cars, they do not detr the core description of the bicycle wheel. The presence successfully fulfills the 'Object(s)' condition by}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b95ac152e88fbc9f73002536cbe3d21200fdd68a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c08b8951b6e39b2e1d6d67f4fe9cb36f3ffe5d9cdb5d140ed3b4527e2eaa58d4
+size 198871
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..ea993f486f5d280de9f8a372080cca094d92db72
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two young girls sitting in a park. The first girl is wearing a blue shirt and has her hair pulled back. The second girl is wearing a white shirt and has her hair down. They are both looking at the camera with expressions of surprise or shock. The background of the video shows a grassy area with trees and a house in the distance. The style of the video is candid and informal, capturing a moment of the girls' lives."
+    ],
+    "video_ids": [
+        "RyJ6JM6_Hg0_20_205to396"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two young girls? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two young girls sitting outdoors, which matches the 'Object(s)' condition. Their facial features, hair, and clothing are consistent with being young girls, and there are no elements that contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two young girls sitting side. which ful the descriptionObject(s)' condition. The presence expressions and clothing, and clothing are consistent with the young,, and there are no additional in contradict this description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f5f0c5470ca40b068ba7a1bacfcf7ab5f48e77c2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4c9a85ea317bf1bd1ddf35d1443aef343270794cb94a62235966a10806809190
+size 215043
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..7b59c5a17f4bcfd148fba32d335adad62586e157
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a group of four women sitting on a stage, each wearing colorful, patterned clothing. The woman on the far left is wearing glasses and has a surprised expression on her face. The woman in the center is larger and has her eyes closed, seemingly in a state of relaxation or perhaps even sleep. The woman on the far right is wearing a black top and has a serious expression. The woman in the middle is wearing a floral print top and has her eyes open, looking towards the camera. The stage they are sitting on has a blue background and there is a potted plant in the background. The style of the video is a candid, unscripted moment captured on stage, possibly during a live event or performance."
+    ],
+    "video_ids": [
+        "8_zR4OppJ-A_33_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Four women sitting on a stage, each wearing colorful, patterned clothing.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows four women seated on a stage, and each is wearing colorful, patterned clothing. The description is accurately fulfilled without any contradictory elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows four women sitting on a stage, each they of wearing colorful, patterned clothing. The women is largely represented by any contradictions elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1d5a2b261c150b3cd0564165556d5729a3081127
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4dcf0f4d89e2906610c212715a1417f53db0fc77831f3a050ec5bbf04ae4f3f7
+size 236136
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..d119d7390969ede86ce1f96340a3551d9a446d2a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young girl with blonde hair, wearing a polka dot shirt and a pink hair accessory. She is standing in a room with a blurred background, which suggests a casual indoor setting. The girl appears to be engaged in a conversation or an activity, as she is looking up and to the side, with a slight smile on her face. In the foreground, there is a book with the title \"Wenn die Katze nicht kommt\" visible, which translates to \"If the cat doesn't come\" in English. The style of the video is a close-up shot, focusing on the girl and the book, with a shallow depth of field that blurs the background. The lighting in the room is soft and natural, creating a warm and inviting atmosphere. The overall impression is that of a candid moment captured in a home or a comfortable indoor space."
+    ],
+    "video_ids": [
+        "6y1lHGSuDp0_15_0to156"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Young girl with blonde hair, wearing a polka dot shirt and a pink hair accessory; a book titled 'Wenn die Katze nicht kommt'.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young girl with blonde hair, wearing a polka dot shirt and a pink hair accessory, which matches the description. The book held by the interviewer is partially visible and appears to have the title 'Wenn die Katze nicht kommt', which aligns with the specified object. The presence of an interviewer and background elements does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young girl with blonde hair, wearing a polka dot shirt and a pink hair accessory, which matches the description. The presence titled by the girl is partially visible and appears to be the title 'Wenn die Katze nicht kommt', which aligns with the description object. The presence of the interviewer and the elements does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4d409fedfe909006913ae9a35cc2d4b159fc44fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c15ceb3670e6be96fde7ebbc05b3631c521094c686d5579afd5fd9514dd78052
+size 224390
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..5fad0fe8482905751c8f493f30a3497ca761143e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man in a pink shirt is seen interacting with a black and tan dog. The man is holding the dog in his arms, and the dog appears to be enjoying the attention. The man and the dog are in a room with a refrigerator in the background. The man seems to be speaking to the dog, and the dog is looking up at him with a happy expression. The overall atmosphere of the video is warm and friendly."
+    ],
+    "video_ids": [
+        "KGZyK3AVyQ4_70_0to151"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a pink shirt and a black and tan dog.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a pink shirt holding a black and tan dog, which matches the core description. The background elements, such as the glass block window and appliances, do not contradict the presence of the man and the dog.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two man in a pink shirt holding a black and tan dog. which matches the description description. The presence and, such as the refrigerator door and and the, do not contradict the main of the man and the dog.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..10ea2c97b8ec593ce2f7494f95a0416fc715fd1d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dc645dcd685f99860d1e9c07f979ef7402dee9cc460417aa979d29bb25be94a2
+size 211249
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..96ed1e93eeb585c206d0a0f326e831be9d195ea4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a bartender in a well-stocked bar, preparing a drink. The bartender, dressed in a blue shirt, is seen in three frames: first, he is reaching for a bottle of alcohol; in the second frame, he is pouring the drink into a glass; and in the third frame, he is holding the finished drink. The bar is filled with various bottles of alcohol, and there are also some fruits, such as limes, visible. The bartender's actions suggest that he is in the process of making a cocktail. The overall style of the video is casual and relaxed, capturing the everyday work of a bartender in a busy bar."
+    ],
+    "video_ids": [
+        "fyZ4bwuPuiI_9_63to228"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bartender, bottles of alcohol, limes, glass? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a bartender standing behind a bar, surrounded by bottles of alcohol, limes, and a glass (a cocktail glass with a lime garnish). All core elements described in the 'Object(s)' condition are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a bartender in behind a bar, which by bottles of alcohol and andimes, and a glass.which cocktail sh). a lemon wedgeish). The the elements of in the 'Object(s)' condition are present and accurately depicted in the video.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f3dedac67e3c6e28ea284dda9292de6c358e906a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1178-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52531f1ece669a1e4167474e9b0589d27535b96e0adb7b436850dd2baf6258ac
+size 149109
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..8c4065147ea525027ee680bf5b091e70dbb58bec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man named Dr. R. Wade Weaver, a Florida cancer specialist, who is standing in a room with a window and blinds. He is wearing a white shirt and a pink tie. The video is a news segment from ABC News, as indicated by the logo in the corner. The style of the video is informative, with a focus on the man and his profession. The room appears to be a typical office or waiting room, with a window that lets in natural light. The blinds are partially closed, suggesting that the room is well-lit but not overly bright. The man is standing in front of the window, which provides a backdrop for the interview. The overall tone of the video is serious and professional, reflecting the gravity of the subject matter."
+    ],
+    "video_ids": [
+        "KYpIcapvNIM_3_0to182"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dr. R. Wade Weaver (a man), window, blinds, ABC News logo? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing Dr. R. Wade Weaver (a man), a window with blinds in the background, and the ABC News logo in the lower right corner. All specified elements are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showing showing a. R. Wade Weaver,a man), a window with blinds, the background, and the ABC News logo in the corner right corner of The these elements are present and do depicted.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..14d0e39f8fb4431cc86a3dbf2c85171b48ad899d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eff1c9af2521a710ee6a0a698cf8dbb18a17646cc983ffea2adfd62b1a0d8122
+size 119949
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..3a7fab3712f013892017da43c291bab751dfad68
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a woman in a bookstore, smiling and looking to her right. She is wearing a green scarf and a black jacket. The bookstore has shelves filled with books and toys. The woman appears to be enjoying her time in the store. The video is a casual, everyday scene, capturing a moment of joy in a bookstore."
+    ],
+    "video_ids": [
+        "AITR0jw80lI_6_0to103"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a green scarf, a black jacket? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman wearing a green scarf and a black jacket, which matches the specified 'Object(s)' condition. The background and other elements do not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a woman wearing a green scarf and a black jacket. which ful the ' objectsObject(s)' condition. The background, additional elements, not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b1ea22268f72790d4a8a3803797bee93a62bb93b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac96a51ad78a7fc9cc33e1fd5f2857011f9be5503f42dbd0edcf75e2da186038
+size 163307
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..bc610f1714e2820914eefc39e750123494ee2080
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman with long curly hair and glasses is seen sitting at a desk with a microphone in front of her. She is wearing a black blazer and a white shirt. The desk is cluttered with various items, including a potted plant, a vase, and a picture frame. The room has large windows that let in natural light, and there are plants placed around the room. The woman appears to be in the middle of a conversation or recording a podcast. The overall style of the video is casual and informal, with a focus on the woman and her surroundings."
+    ],
+    "video_ids": [
+        "1J4jLTZpl4M_2_54to180"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with long curly hair and glasses, a desk, a microphone, a potted plant, a vase, a picture frame, a woman wearing a black blazer and a white shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with long curly hair and glasses, sitting at a desk with a microphone in front of her. There are potted plants and a picture frame visible in the background. She is wearing a black blazer and a white shirt, matching the description. Additional elements like a vase and an inset image of two men do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a woman with long curly hair and glasses, sitting at a desk with a microphone, front of her. There are potted plants and a vase frame on on the scene, The is wearing a black blazer and a white shirt, which the description. The elements like the vase are a object image are a people are not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7013dbdc1095ef0b3bb1f86ea030c3efa709b9b4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5e8a44c01b45250da27354f02c8b2759a37fc037f3c1e3d7853ef56e6e55118
+size 132661
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..eed7e5124584e0c5bba172921b23f1cd4c4a7aa1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of an elephant's face as it is being sprayed with water. The elephant's skin is visibly wet, glistening under the spray, highlighting its textured surface. The water droplets are captured mid-air, creating a dynamic and refreshing scene. The background is a lush green, suggesting a natural outdoor setting, possibly a sanctuary or a park. The elephant's eyes are partially closed, indicating a sense of relaxation or enjoyment from the water spray. The camera remains steady throughout the sequence, focusing on the elephant's face and the water droplets, providing a detailed and intimate portrayal of the moment. The overall atmosphere is serene and peaceful, emphasizing the connection between the animal and its environment."
+    ],
+    "video_ids": [
+        "6ae6e1142f7f36f9cc2adecc5f9540c9b50b64248f8d8b0db14a070412b40af0"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An elephant's face, water droplets in mid-air? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a close-up of an elephant's face, with visible water droplets in mid-air spraying across its skin. The focus remains on the elephant and the water, fulfilling the described conditions without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts an close-up of an elephant's face, with water water droplets in mid-air, from the skin. The presence is on the elephant's the water dro fulfilling the ' ' without any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..546b50a00f34257ac220de8517fbfa12834b5015
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5bc4ff31d6719417abdf89d90fea20f157dbd651ee1f214fe3ca1f043d398ab4
+size 355191
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..2620090a38dedf00acba5e4cbecb0058e31c65c6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a black vest standing in a room with a green wall. He is gesturing with his hands, possibly explaining something. Behind him, there is a shelf with various objects, including a chess set, books, and figurines. The room has a vintage feel, with a reel-to-reel tape recorder and a film projector on a table. The man appears to be in the middle of a conversation or presentation, as he is actively engaging with the viewer. The overall style of the video is casual and informative, with a focus on the man and his surroundings."
+    ],
+    "video_ids": [
+        "2wRcDM_5rjs_9_28to217"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black vest, a shelf with a chess set, books, and figurines, a reel-to-reel tape recorder, and a film projector.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black vest, standing in front of a shelf that contains a chess set, books, and figurines. A reel-to-reel tape recorder and a film projector are also visible on a table behind him. All specified objects are present and do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man in a black vest, standing in front of a shelf that contains a chess set, books, and figurines. There reel-to-reel tape recorder and a film projector are also visible on the table in him. The these elements are present and match not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bb525895c77e5ef41d69fdbb8daef600ceb313f9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd9478fe54d778137a613194ce04ae7235d812898f72581278c5f7af68044add
+size 154544
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..f32c608920bc60bbdbe8f78d4e67a80df04a27d7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a kitchen scene where a yellow mixer is being used to mix ingredients. The mixer is placed on a wooden table, and there are several bowls and spoons around it. The first bowl contains butter, the second contains flour, and the third contains sugar. The mixer is in the process of mixing the ingredients together. The style of the video is a simple, straightforward cooking tutorial, focusing on the process of mixing ingredients for a recipe. The video is likely to be used for instructional purposes, teaching viewers how to prepare a specific dish."
+    ],
+    "video_ids": [
+        "eZzBgIuLdHA_31_0to113"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Yellow mixer, wooden table, three bowls, spoons? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a yellow mixer, a wooden table, and three bowls (one large bowl with flour, one small bowl with butter, and one small bowl with cream). A spoon is visible in the large flour bowl. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a yellow mixer, a wooden table, and three bowls.one with bowl with the, one smaller bowl with a, and one small bowl with sugar). There spoon is also in the video bowl bowl. The elements elements of in present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a9b13a5035aa2c6dc70a78f25a374b27b07c2ab1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a280351c883b4c15ae9e539b1535400af0af189b3490322bd4624b547ae4d3e
+size 85027
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..7e71a90b6dfc462d76207db697825296663cf487
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a pair of white over-ear headphones on a stand. The headphones are positioned in the center of the image, with the stand providing a stable base. The stand is silver and has a circular base. The headphones are connected to a device, which is not visible in the image. The background of the image is blurred, but it appears to be a room with a desk and a monitor. The style of the video is simple and straightforward, focusing on the headphones and their stand. The image is static, with no movement or action taking place. The headphones are the main subject of the video, and the stand is used to display them. The background is not the focus of the video, and it is not detailed. The image is clear and well-lit, with the headphones and stand being the main points of interest."
+    ],
+    "video_ids": [
+        "Be7VtE5v94U_59_0to129"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A pair of white over-ear headphones and a silver stand.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a pair of white over-ear headphones mounted on a silver stand, which matches the core description. Additional elements like a speaker and a plant in the background do not contradict the main object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a pair of white over-ear headphones resting on a silver stand. which matches the description description provided There elements like the wooden or a keyboard in the background do not contradict the main focus description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a775ca930d49ca7d3d7c2c12be125fd375793d59
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7041999259c718ab77e270754a32090d50a3f7265eb63696ea9263a7e6796b48
+size 59876
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..3d60f0d5090076a965e0d7e0ad43d0f6a84938fa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a tan coat and a green backpack walking down a street. He is wearing glasses and a hat. The street is lined with buildings and there are chairs and tables outside a restaurant. The man is carrying a bottle of water in his backpack. The video is shot in a realistic style with natural lighting."
+    ],
+    "video_ids": [
+        "FoDyr4jcnbU_21_127to289"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a tan coat and green backpack, wearing glasses and a hat, and a bottle of water in the backpack.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a tan coat, a green backpack, glasses, and a hat. A bottle of water is visibly attached to the side of the backpack. These elements are consistent with the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a person wearing a tan coat and a green backpack, and, and a hat. The bottle of water is visible placed to the backpack of the backpack, The elements match consistent with the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..76c23d2a1c4ab916941a93a81b689c764486f683
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1179-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da6e84fad3f865638a53ade297f46faf3d4463686afafe0899d849414319720e
+size 245983
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..08bd5870bad706506f3174a7eec93b6524837b94
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse of a car's interior, showcasing the dashboard and steering wheel. The car is a Mercedes-Benz, as indicated by the logo on the steering wheel. The dashboard is sleek and modern, with a digital display that shows the car's speed and other information. The steering wheel is black and has a red center, adding a touch of color to the otherwise monochrome interior. The car's interior is well-lit, with natural light coming in through the windows. The seats are black and appear to be made of leather. The car is parked in a lot, as evidenced by the other cars visible in the background. The video does not show any people or animals. The style of the video is straightforward and unembellished, focusing solely on the car's interior."
+    ],
+    "video_ids": [
+        "Y_QGqWUsxqg_30_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Mercedes-Benz car, dashboard, steering wheel, seats? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the interior of a Mercedes-Benz car, including the dashboard, steering wheel, and seats. All specified objects are prominently featured and accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows the interior of a Mercedes-Benz car, including the dashboard, steering wheel, and seats. The the objects are present visible, match depicted, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f761981b6d6b79f49198437011ac7d8ec4c9897c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b79eb553e3211ee151c33c89f01ed04833ec04d515e23f647d020d7480e4d43e
+size 253016
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..145666767aa1d3272d5029c9bf75bd8cc460da85
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a cozy outdoor setting during the fall season. A wooden rocking chair is the main focus, positioned against a rustic brick wall. The chair is adorned with a colorful throw pillow, adding a touch of warmth to the scene. On the brick wall, a black lantern and a small pumpkin are placed, creating a charming autumn ambiance. The pumpkin's vibrant orange color contrasts beautifully with the brick wall. The overall style of the video is rustic and inviting, evoking a sense of comfort and tranquility."
+    ],
+    "video_ids": [
+        "caJKHZ26GG0_24_0to134"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Wooden rocking chair, rustic brick wall, colorful throw pillow, black lantern, small pumpkin? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a wooden rocking chair, a rustic brick wall, a colorful throw pillow (with geometric and text designs), a black lantern, and small pumpkins (including orange and white varieties). All specified elements are present and accurately depicted without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as showcasing showing a wooden rocking chair, a rustic brick wall, a colorful throw pillow,with a patterns floralural), a black lantern, and a pumpkins.one one and green ones). The these objects are present and match depicted in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..029c9dd29d275f6090c19cb192c61284f66519a5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59e35331f5dd76ca6e44cb1026e8ce8dc90f6d95cad44c6cd43a11364519117a
+size 178042
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..61c897d0b11dfb7e4ed344950baef01bc7167b92
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a group of six young men standing in front of a large, ornate building with intricate architectural details. They are all smiling and appear to be engaged in a lively conversation. The men are dressed in casual attire, with a mix of colors and patterns. One man is wearing a pink shirt, another is in a black jacket, and a third is in a white and black striped shirt. They are all standing close to each other, suggesting a sense of camaraderie and friendship. The building behind them is quite grand, with large windows and a prominent entrance. The overall style of the video is casual and friendly, capturing a moment of joy and connection among the group of friends."
+    ],
+    "video_ids": [
+        "8HuTRGDaXPc_20_0to120"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Six young men, one large ornate building? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows six young men standing and interacting in front of a large, ornate building that matches the description. The building features Gothic-style windows and stone architecture, consistent with a historic university setting. The subjects are accurately represented as young men, and their presence and interaction align with the 'Object(s)' condition. No elements contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts six young men standing in smiling in front of a large orn ornate building. matches the description. The building has architectural architectural architecture and decorative architecture, which with the large or or. The presence are positioned depicted as young men, and the casual in the fulfill with the 'Object(s)' condition.\"\n There additional contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0a4551310fea9d63a9472cafa502771af061f4b4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:01b5acff0969095bbdb9b30021db8e0ea76aa83b9f6a32f343ecb6c68ae56cad
+size 197615
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..1dc2af8a8fca2608f0fb90f3f8257d53365df626
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are engaged in a conversation. The man is wearing a blue shirt and has a microphone attached to his shirt, suggesting that he might be a reporter or an interviewer. The woman is wearing a white blouse and has long brown hair. They are standing in front of a green hedge and a fence, which indicates that they might be in a park or a garden. The background also includes a bench and a tree, adding to the natural setting. The man is looking at the woman with a smile on his face, indicating a friendly and positive interaction between them. The overall style of the video is casual and informal, with a focus on the conversation between the two individuals."
+    ],
+    "video_ids": [
+        "O6eN5qFchRU_8_0to123"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman sitting together, engaging in conversation. Their presence and interaction fulfill the 'Object(s): A man and a woman' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man and a woman standing together, which in what. The attire and interaction fulfill the 'Object(s)' A man and a woman' condition. any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6306bb1728644353011358430aad670476431439
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fc4d3ee5db69df97d874870f28bfdc7509bbfd00da6f84ff6427002f0f36d20
+size 214186
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..a0537538dee86cb2d7b4460da49000b348919ad1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a close-up of a blonde doll with large blue eyes and a pink bow in her hair. She is wearing a pink top and is sitting on a carpeted floor. The doll appears to be in motion, as if she is moving her mouth and eyes, giving the impression of a lifelike expression. The background is blurred, but it seems to be an indoor setting with a door and a blue wall. The style of the video is realistic, with attention to detail in the doll's facial features and clothing. The focus is on the doll, making her the main subject of the video."
+    ],
+    "video_ids": [
+        "Kucf3JBMeRY_2_0to199"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blonde doll with large blue eyes and a pink bow in her hair. She is wearing a pink top and sitting on the carpeted floor.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a blonde doll with large blue eyes and a pink bow in her hair, wearing a pink top and sitting on a carpeted floor. All key elements of the description are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a blonde doll with large blue eyes and a pink bow in her hair. wearing a pink top. sitting on a carpeted floor. The the elements of the description are present, match depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fcd8825bb40de0c8d465aa38fd1937cded7c201a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1d20bf76b72d2c79041817595b08a50b67eeb7ad8363ec46d8e9e9c4aa5ff20
+size 106505
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..89fb9492d4e9a191cad1310cfc7a46e1696cd711
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person using a yellow and black power saw to cut through a piece of metal. The saw is held firmly in the person's hand, and the blade is spinning rapidly. The metal being cut is placed on a sturdy workbench, and the person appears to be concentrating on their task. The background of the video shows a parked car and a leaf-covered lawn, suggesting that the scene is taking place outdoors. The style of the video is straightforward and documentary, focusing on the action of the saw cutting through the metal."
+    ],
+    "video_ids": [
+        "-FP3MmF0GcI_16_0to190"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a yellow and black power saw, and a piece of metal on a workbench.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person interacting with a yellow and black DeWalt power saw, which is mounted on a workbench. Although the saw is not actively cutting metal, the person is handling the saw blade and the setup suggests metal cutting is intended or in progress. The presence of a metal workbench surface is also evident. The description does not require the metal to be actively cut, only to be present on the workbench, which is satisfied.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person using with a yellow and black powerWalt power saw, which is cutting on a piecebench. The the video is not actively cutting in in the presence's holding the saw,, appears work suggests theworking is the. has progress. The presence of a piece workbench and and also visible, The video of not specify the saw to be actively being, only that be on on the workbench, which is the by}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..50db4949bdc7449635505d06ba4bf36008c3b448
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:16bd7807c6adb83d52389d126aee628a205bc28adeaa4c3844540760d8b5f869
+size 108328
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..796575b979a9fe2a0415035a0d7c7446d56e876e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man in a blue shirt and a red and white baseball cap standing in a field with a mountainous landscape in the background. He appears to be in a relaxed posture, possibly taking in the view or waiting for someone. The field is open and flat, with a few scattered bushes and trees. The mountains rise steeply from the ground, their rugged terrain covered in a mix of green and brown vegetation. The sky is clear and blue, suggesting a sunny day. The overall style of the video is naturalistic, with a focus on the man and his surroundings. The camera seems to be positioned at a distance, capturing the man and the landscape in a single frame. The lighting is bright and even, indicating that the video was likely taken during the day."
+    ],
+    "video_ids": [
+        "43TPkqrArjM_8_0to103"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue shirt and a red and white baseball cap.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a blue shirt and a red and white baseball cap, which matches the description. The man is the central focus and his attire is consistently visible throughout the frames, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing a blue shirt and a red and white baseball cap. standing matches the description provided The background is standing central figure of is attire is consistent visible throughout the frames. fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c0bc99641386a2c8345db3785b599fb45ab060f3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af132b114f0ab5ab89a6789fe7368c487686cdae2f2cb9f10fee154fe333ea66
+size 236210
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..830366ba81d5ab265ed3196d99233915999b3b62
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a vintage car engine, focusing on the intricate details of the engine's components. The engine is silver and features a prominent air scoop on top. The video captures the engine from different angles, showcasing the craftsmanship and design of the vintage car. The style of the video is informative and detailed, providing a comprehensive view of the engine's components and their arrangement. The video is likely intended for car enthusiasts or those interested in the mechanics of vintage cars."
+    ],
+    "video_ids": [
+        "AbYvsPDfcv4_5_324to454"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Vintage car engine, engine components? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a vintage car engine with visible engine components such as carburetors, hoses, and metal parts. The engine is detailed and appears to be from a classic car, matching the description. The presence of a logo in the corner does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a close car engine with various engine components such as theuretors, hoses, and mechanical parts. The style appears depicted and appears to be from an classic vehicle, which the description of There focus of additional single or the corner does not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fd11d4d82330621e34e3c834771bfbe1ddc8c8de
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-118-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3175d83bcb66bbe435206d5d1f3c17057e18e7e00ed7d3b61166afc6d1a038a6
+size 208347
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..3a0db3ed3665f6c4d33aa0dc5e55d49c6bedea2d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen holding a pink tutu dress with a silver bow on it. She is standing in front of a window, which is letting in natural light. The woman is wearing a white sweater and a necklace. The video is likely a tutorial or a review of the tutu dress, as the woman is examining the dress closely. The style of the video is casual and informative, with a focus on the tutu dress."
+    ],
+    "video_ids": [
+        "7qZ2gbOm20E_0_26to185"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a pink tutu dress with a silver bow, a white sweater, and a necklace.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman holding a pink tutu dress with a silver bow, wearing a white sweater and a necklace. These elements match the description provided in the 'Object(s)' condition, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman holding a pink tutu dress with a silver bow. wearing a white sweater, a necklace. The elements match the description provided, the questionObject(s)' condition. indicating there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2717914fc0f3da0bdc0244b37cc63c33cd4ac62c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8e49a00ab59bc7f3c1ecb8b0f53d4688828afc6dd9aad0daa4ee6adbe2a1e967
+size 195497
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..2eee36ebb951d3c395bdada9f0d58569b6f171c6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a luxury bus, which is spacious and well-appointed. The bus features a high ceiling with a large skylight, allowing natural light to flood the interior. The seating area is arranged with two rows of comfortable, leather seats, each with its own headrest and armrest. A small table is situated between the two rows, providing a convenient surface for passengers to place their belongings or enjoy a meal. The floor of the bus is made of polished wood, adding a touch of elegance to the overall design. The bus is equipped with large windows that offer a view of the outside world, and the curtains can be drawn for privacy. The overall style of the video is sleek and modern, with a focus on the luxurious amenities and design of the bus."
+    ],
+    "video_ids": [
+        "6U9PiwJ5VGg_9_98to232"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Leather seats, headrests, armrests, a small table, curtains? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows leather seats with headrests and armrests, a small table (a bar-style table), and curtains covering the rear window. All specified elements are present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a seats, headrests and armrests. a small table inwhich desk table table in and curtains on the windows windows. The the objects are present, match depicted, the image,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..af4a4f269e0b5590d639080c9efe8336768710aa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:39bf1ec6d48e4753217cde68455699690fc16d711e0504c10782a669033daf2b
+size 99806
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..b92c3886167d5b02e83ead9fb6a9372c1b6cacc6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment from CTV News Channel, featuring two individuals in a studio setting. The first individual is a man with gray hair, wearing glasses, a suit, and a tie. He is seated in front of a backdrop that shows a cityscape with a body of water. The second individual is a woman with short hair, wearing a dark dress with a light-colored top. She is seated in front of a backdrop that shows a studio with multiple monitors and a news ticker. Both individuals are engaged in a conversation, likely discussing a topic of current interest. The style of the video is professional and informative, typical of a news broadcast."
+    ],
+    "video_ids": [
+        "AL8YG3LdS2M_10_26to211"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two individuals - a man with gray hair and glasses, wearing a suit and tie; a woman with short hair, wearing a dark dress with a light-colored top.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two individuals as described: a man with gray hair and glasses, wearing a suit and tie, and a woman with short hair wearing a dark dress. The woman's attire is consistent with a dark dress, and while the description mentions a 'light-colored top,' it is likely referring to a light-colored undershirt or collar, which is not clearly visible. The core description is largely accurate.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two individuals, described: a man with gray hair and glasses wearing wearing a suit and tie, and a woman with short hair, a dark dress with The setting's dress includes a with a light dress with but the the description does a lightlight-colored top,' the is not that to the white-colored blousehirt or blouse, which is visible explicitly visible in The presence elements of largely fulfilled.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a1dacc6af311b73180d1cc6bc92df33cb75c655b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c90c575af701b66cb8cad6686ce7d790afa86ef17b148366bac2af7b3a64b100
+size 131527
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..5e698c88494d00688be17131ad40a99149eaa1b9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a black shirt, who appears to be engaged in a conversation. The man is looking to his left, suggesting that he is interacting with someone off-camera. The background is dark, which puts the focus on the man. The style of the video is a close-up shot, which emphasizes the man's facial expressions and body language. The lighting is dim, creating a moody atmosphere. The man's expression is serious, indicating that the conversation might be of importance. The overall style of the video suggests a dramatic or intense scene."
+    ],
+    "video_ids": [
+        "tmFqmYtDHkE_15_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a black shirt, which matches the 'Object(s)' condition. The background is dark, and the focus is on the man, with no conflicting elements present that would contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a man wearing a black shirt. which ful the descriptionObject(s)' condition. The focus is dark, and the focus is on the man, making no additional elements present.\"\n would contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8ba156302e4b79b7d998602a9da2681d3e3d087f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cfdfd5b10185829cc8b42aaaab85602709cea62f9993c917df4123280f861be5
+size 191739
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..d7aa5a1b4447855b7cb9ace68f59ec441e8b9815
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a white sedan car parked on a road with a scenic mountain backdrop. The car is captured from a rear three-quarter angle, highlighting its design and features. The car's taillights are illuminated, suggesting that the car is either in motion or has been recently driven. The car's design includes alloy wheels and a distinctive rear bumper. The road appears to be a two-lane highway, and the surrounding environment suggests a rural or semi-rural setting. The video is likely a promotional or review video for the car, showcasing its design and performance in a picturesque setting."
+    ],
+    "video_ids": [
+        "XiC2kbmjMWY_12_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white sedan car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a white sedan car, which is the central object. The car is clearly visible throughout the video, and its features such as the body shape, color, and wheels are consistent with a white sedan. The background elements (mountains, road, water) do not contradict the description of the object.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a white sedan car, which is the main object in The car is depicted visible in the frames, and its design, as the ta,, wheels, and design are consistent with the sedan sedan. The presence and,mountains, trees, and reflections do not contradict the description of the car but}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a0bc98568138091141a64c75ed952fd285616497
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c62a3bba90488d34ee2871e34644b989ee53925f24e6dc18402e7f4d71e0eb1d
+size 217895
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..995d7bc589f4081f488a37940cc6b3e9655da72d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with long, dark hair and a white blouse with black accents. She is seated and appears to be engaged in a conversation. The setting is a studio with a backdrop that includes a floral arrangement and a decorative wall. The lighting is bright and even, highlighting the woman's features and the details of her attire. The style of the video suggests it could be a segment from a talk show or a similar television program."
+    ],
+    "video_ids": [
+        "VociqUXNI8s_2_22to230"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with long, dark hair and a white blouse with black accents.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The woman in the video has long, dark hair and is wearing a white blouse with black accents, which matches the description. The presence of other elements, such as the background and another person partially visible, does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video in the video has long, dark hair and is wearing a white blouse with black accents, which matches the description provided The setting of a elements such such as the background and the person, visible, does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..26ad27be9bd97d138dd77dedc97ad021ec3f51f3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1c78aa8ac20012297e83c41b4a9492600a584dd7c9313186b809292756ed46b0
+size 118828
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..d7e6e90e9c4b5e4f1ab4ebee6f8cd06d0800594f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen in a workshop setting, holding a large, round, woven basket on his head. He is wearing a black t-shirt with the number 4 printed on it. The workshop is filled with various objects, including shelves stocked with items, a rope hanging on the wall, and a potted plant. The man appears to be in the middle of an activity, possibly related to the items in the workshop. The video captures a moment of everyday life, with the man and his surroundings providing a glimpse into his world."
+    ],
+    "video_ids": [
+        "KsgeeQKQviA_21_111to275"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a large round woven basket, a black t-shirt with the number 4 printed on it.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man holding a large round woven basket in front of his face, and he is wearing a black t-shirt with the number 4 printed on it. These elements are clearly visible and match the description. The background and other objects do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a large round woven basket on a of his head, which he is wearing a black t-shirt with the number 4 printed on it. The elements match consistent present and match the description provided The background includes additional objects in not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2f14f2494647be165fdb2057a77828c385bab784
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a513cadef5910fb6faf5359e179b93ef2b7ccdf3f2647f88d1b8b01742afbce7
+size 167775
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..ea0c3225cfa250b651487a0f16fe4cdadb514e65
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a friendly interaction between two men in a modern, well-lit room. The man on the left is wearing glasses, a red shirt, and a blue hoodie, while the man on the right is dressed in a black leather jacket and a dark shirt. They are shaking hands, indicating a positive exchange or agreement. The room has a contemporary design with blue and red lights hanging from the ceiling, creating a warm and inviting atmosphere. The overall style of the video is casual and professional, capturing a moment of connection between the two individuals."
+    ],
+    "video_ids": [
+        "5rmnLJ7GSMM_32_0to103"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men standing and interacting with each other, fulfilling the 'Object(s): Two men' condition. Their presence and interaction are the central focus of the video, with no elements contradicting this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two men engaging and shaking with each other. which the 'Object(s)' Two men' condition. There attire and interaction are the central focus of the video, and no additional thating this description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6d086ab5157937d6d2f33818d889966b31c1e6c0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1180-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:585640554983ac71e14afdb2e3e76dc1008252809b62f64ecaef931697024275
+size 183289
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..14303461e2ed2398640a53c2270a6c0056b3f445
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the majestic view of a cityscape from a high vantage point, focusing on a large, white, pointed building with a green roof. The building stands out against the backdrop of the city, its pointed roof reaching towards the sky. The city below is a mix of buildings, trees, and water, creating a vibrant and dynamic scene. The video is taken during the day, with the sun shining brightly, casting a warm glow on the buildings and the water. The overall style of the video is aerial, providing a bird's eye view of the city and the building, emphasizing the grandeur and scale of the architecture."
+    ],
+    "video_ids": [
+        "XBzpnGEAtLY_17_239to395"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Large, white, pointed building with a green roof? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large, white, pointed building with a green roof, matching the description. The structure is a prominent tower with a pointed spire and green-topped finials, set against a cityscape. The core description is accurately represented, even though the surrounding urban environment and distant landscape are also visible.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a large, white, pointed building with a green roof, which the description provided The building is prominent prominent feature in a distinctive topire, a roofingint sectionial, which against a backdropscape and The presence elements is accurately represented in and though the video elements environment and water water are also visible,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5b76c837780ff8bfae5644786bdfe2b31d82720e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2adcca1e2691d9323f2f7f5df60bf102535f9a166d0baddb91a39a3756dbbc98
+size 202047
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..37b36b4f29ee69adecc2120be21cb878792b9bde
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a warm family gathering around a dinner table. The family members, including a man, a woman, and a child, are seen preparing and serving food. The table is filled with various dishes, including a large bowl of soup, a plate of salad, and a tray of sandwiches. The family members are engaged in conversation, creating a sense of togetherness and warmth. The setting appears to be a cozy kitchen, with a window allowing natural light to fill the room. The overall style of the video is candid and intimate, capturing a moment of family bonding and shared mealtime."
+    ],
+    "video_ids": [
+        "0gH5HI89Syc_99_0to138"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a woman, a child, a large bowl of soup, a plate of salad, a tray of sandwiches.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple people including a man, a woman, and a child gathered around a table. There is a large metal pot (likely containing soup) and a plate with salad and sandwiches visible on the table. These elements match the 'Object(s)' condition described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a people around a man, a woman, and a child, around a table. There is a large bowl bowl thatwhich containing soup) on a plate of salad. a.. the table. The elements match the descriptionObject(s)' condition described in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..aa38f4dd9096aedf787657bba8bfedfd2cd57819
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e291c0b7a98a538ca2ec89847c882f81ccc7da6466b1ad9f23e852ce24fbf452
+size 164336
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..b60eecd39a0e2956ea4f888ab1dbebc0d7f2b5fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a group of people exploring a large indoor train museum. The museum is filled with several white and blue bullet trains, each with a sleek and modern design. The trains are parked in a row, allowing visitors to walk between them and admire their details. The people in the video are seen walking around, taking pictures, and discussing the trains. The museum itself has a high ceiling and large windows, allowing natural light to flood in and illuminate the exhibits. The overall atmosphere of the video is one of fascination and curiosity, as the visitors immerse themselves in the world of high-speed rail travel."
+    ],
+    "video_ids": [
+        "QkvtVw2u6NM_102_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Several white and blue bullet trains with sleek and modern designs.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows several white and blue bullet trains with sleek and modern designs, which matches the description. The trains are prominently displayed in a large indoor space, and their streamlined, aerodynamic shapes are clearly visible, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts several white and blue bullet trains with sleek and modern designs. which align the description provided The trains have positioned displayed in a high, setting, likely their design shapes futuristicodynamic shapes and consistent visible, indicating the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..149af59086672991109b8b50af736aad370556f8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b627393daaa5f2c635e99b31142c3369d94000890f71bca31fdf18b2f5dfb55
+size 262157
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..6cda6bff990dd31e2d3e3a813929aa20885b90ea
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a colorful and whimsical animation featuring a windmill, a lighthouse, and three balloons. The windmill, with its vibrant red and yellow colors, is situated on a green hillside. The lighthouse, painted in shades of blue and white, stands tall in the background. The three balloons, each in a different color (red, blue, and green), are floating in the sky above the windmill. The scene is set against a backdrop of a clear blue sky dotted with fluffy white clouds. The overall style of the video is playful and cartoonish, with a focus on bright colors and simple shapes."
+    ],
+    "video_ids": [
+        "bJi26_HWRwE_57_0to129"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Windmill, lighthouse, three balloons? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a windmill on the left, a lighthouse in the center, and three balloons (red, blue, green) tied to the lighthouse. These core elements are present and match the description, even though additional balloons and other objects are also visible.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a windmill, the left side a lighthouse on the background-right and three balloons floatingred, blue, and) floating to the windighthouse. The elements elements match present and match the description. fulfilling though the elements are a elements are also visible in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6e5bac95ab27d12d7fabbc56a92e0db838f97a5e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:15b18eb28ecbf1fe0a286c87143254dd2d960cdeb626539ec3072e9b479ab06f
+size 195544
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..984f4aa339786ee421428a79f98b84bb78f1afb6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen in a kitchen, preparing a cake. The cake is placed on a table, and the person is adding chocolate cookies to the top of the cake. The person is using a knife to place the cookies, and the cake is covered in chocolate frosting. The kitchen has a refrigerator in the background, and the person is wearing an apron. The video captures the process of decorating the cake, from the initial placement of the cookies to the final result. The person's focus and attention to detail are evident in their actions. The overall style of the video is a simple, straightforward depiction of a cake-decorating process, with no additional embellishments or distractions."
+    ],
+    "video_ids": [
+        "A0SnAKVw76Y_91_141to267"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person, cake, chocolate cookies, knife, refrigerator? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a person decorating a cake with chocolate cookies, and a refrigerator is visible in the background. Although a knife is not visibly present, the core objects mentioned (person, cake, chocolate cookies, refrigerator) are all present and correctly depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a person ( a cake with chocolate cookies and using the knife is visible in the background. The the knife is present explicitly used in the action elements ( inPerson, cake, chocolate cookies, refrigerator) are all present and correctly described in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..74275bdd81a4b0a43d59875468c4f368fe8927b0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ef64c0f62641412868811afab55525150e4d37ff34f27afbb350e6f1c2acdfa
+size 161542
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..c0f10aeaa2ff3baae375d888895c84229a922d9e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the journey of a blue and white propeller airplane from takeoff to landing. The airplane, with its vibrant blue and white color scheme, is the main focus of the video. The propeller, a key component of the airplane, is prominently featured in the foreground of each frame.  In the first frame, the airplane is captured in mid-flight, soaring through the clear blue sky. The propeller is in motion, indicating the airplane's forward movement.  In the second frame, the airplane is seen flying over a lush green field. The propeller continues to spin, maintaining the airplane's speed and direction.  In the third frame, the airplane is preparing for landing. The propeller slows down, signaling the airplane's descent. The airplane is now flying over a runway, ready to touch down.  Throughout the video, the airplane's position and the propeller's motion provide a sense of progression and movement. The clear blue sky and the green field add to the overall aesthetic of the video, creating a visually appealing and engaging experience for the viewer."
+    ],
+    "video_ids": [
+        "r26RCI69kUk_40_0to120"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Blue and white propeller airplane, propeller? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a blue and white propeller airplane with a clearly visible propeller. The aircraft's engine and propeller blades are detailed and central to the shot, matching the description. Additional elements like the background scenery and scaffolding do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a blue and white propeller airplane with a visible visible propeller. The airplane is design is propeller are are in and in to the frames, fulfilling the description of The elements like the runway and do theolding do not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f9f083e92cdb72de604495d813488c47a648a664
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0f5d6678c089da9dd421ea5c15a4708212c77082bb618b2825a76434745bfac
+size 303980
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..9a412ce9da8f8c94a99c4873cb20986b31be8adf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a professional setting, possibly a meeting or a conference. She is wearing glasses and a white shirt, and she is holding a blue pen in her mouth. She appears to be deep in thought or contemplating something. The background shows a whiteboard with some writing on it, suggesting that the meeting is taking place in a conference room. There are other people in the room, but they are not the focus of the video. The woman's thoughtful expression and the pen in her mouth suggest that she is actively engaged in the meeting and is likely taking notes or preparing to speak. The overall style of the video is realistic and professional, capturing a common scene in a business or academic setting."
+    ],
+    "video_ids": [
+        "Risf_QQGX2o_40_0to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a blue pen, a whiteboard? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman wearing glasses, holding a blue pen, and sitting in front of a whiteboard. The background also includes other people interacting with the whiteboard, which aligns with the core objects mentioned in the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a woman wearing a and holding a blue pen, and there in front of a whiteboard. The presence and includes a people,, the whiteboard, which iss with the presence description mentioned in the description.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5c8ca5a63e346c43cac3b18080e41de8ae583034
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:752a4972f6028f59052851278f4cd04caff95f92a5f898caed743ab20053e148
+size 174630
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..8db7a58f421b179e0eb1d6f97e0b622f94089e58
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a thrilling moment in a football game. The scene is set in a stadium filled with spectators, their faces a blur of anticipation and excitement. The main focus is on two players, one from the Cowboys and the other from the Redskins, both in the midst of an intense play.  In the first frame, the Cowboys player, donned in a white jersey, is seen running towards the end zone, his eyes focused on the goal. His body language suggests a sense of urgency and determination.  The second frame captures the moment when the Redskins player, wearing a maroon jersey, leaps into the air, his arms outstretched to intercept the ball. His athletic form is a testament to the agility and speed required in the sport.  The third frame shows the aftermath of the play. The Redskins player, having successfully intercepted the ball, is seen holding it securely in his hands. The Cowboys player, on the other hand, is seen looking up at the sky, perhaps in disbelief or frustration at the turn of events.  The video is a dynamic snapshot of a football game, capturing the high-stakes action and the split-second decisions that define the sport."
+    ],
+    "video_ids": [
+        "1oVL5g32TKw_52_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two football players - one in a white jersey (Cowboys) and one in a maroon jersey (Redskins).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two football players: one in a white jersey with blue accents (Dallas Cowboys) and another in a maroon jersey with yellow accents (Washington Redskins). Both players are clearly visible and identifiable by their team colors and uniforms, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two football players, one in a white jersey and the and,Cow Cowboys) and one in a maroon jersey with white accents (Washington Redskins). Both players are on visible, are by their uniforms colors and numbers. fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..924076ad832aac08e1d58e66ee26345905725a7e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1181-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:875412d6a62bcf33d3402219f6874d66f9208c15f629e59b5b821ca7d585470e
+size 297642
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..d2763c09cc5bf3d313da2bed962ad9840bfac350
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a drone shot of a luxurious estate with a large pool and a green lawn. The estate is surrounded by a desert landscape with mountains in the background. The pool has a unique design with multiple water features and is surrounded by a well-manicured lawn. The house itself is large and has a grand entrance with a circular driveway. The video is likely meant to showcase the estate's amenities and the surrounding landscape. The style of the video is aerial and it provides a comprehensive view of the property and its surroundings."
+    ],
+    "video_ids": [
+        "e5oDoVn5pyo_6_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Large pool, green lawn, desert landscape, mountains, house with a grand entrance and circular driveway.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a large pool with winding water features, a well-maintained green lawn, a desert landscape with sparse vegetation, distant mountains, and a grand house with a prominent entrance and circular driveway. All core elements described are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as showcasing showing a large pool, various edges features, a green-maintained green lawn, a desert landscape with distant vegetation, and mountains, and a house house with a circular entrance and circular driveway. The elements elements of in present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9f3a5ba19eca547ec4b190d1540cdda06089de09
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ae22c6fe5fc4835719774a026afb6fec21a3ba52fcfaeff1b9af3d747288b90
+size 260398
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..64b81aae5ec9e2baa87c76907f26433c9506ca25
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman with long blonde hair is seen in a cozy room, enjoying a glass of red wine. She is dressed in a gray sweater and holds the wine glass in her right hand. The room is warmly lit by a fireplace, which is located to her left. The woman appears to be in a relaxed and comfortable state, savoring her drink. The overall atmosphere of the video is calm and inviting, with the woman's enjoyment of her wine being the central focus."
+    ],
+    "video_ids": [
+        "Wyu5A3igIbI_108_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with long blonde hair, a gray sweater, and a glass of red wine.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with long blonde hair, wearing a gray sweater, holding a glass of red wine. These elements are consistent with the description, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with long blonde hair wearing wearing a gray sweater, and a glass of red wine. The elements match consistent with the description provided and there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..611aae50fdf0a91fcd4f40efdf6f23e0c4f574fb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e158776637a2939dc3751131c18a3218f9d7d420e7cd65be10fe034839c7a12
+size 188271
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..634249f11b0f21b62f7973a821eb4d9fba68424e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car, focusing on the front passenger seat. The seat is upholstered in a light brown leather with a diamond-stitched pattern. The seat is positioned in a way that the backrest is facing the camera, and the headrest is visible. The seat is empty, and there are no passengers in the car. The car's interior is well-lit, with natural light coming in from the windows. The style of the video is a close-up shot, focusing on the details of the seat and the surrounding interior. The video does not contain any text or additional elements. The overall impression is of a luxurious and well-maintained vehicle."
+    ],
+    "video_ids": [
+        "oozvOIje5Rg_196_0to133"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Front passenger seat upholstered in light brown leather with a diamond-stitched pattern.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the front passenger seat upholstered in light brown leather with a distinct diamond-stitched pattern, matching the description. The camera pans out to reveal more of the interior, but the core object (front passenger seat) remains consistent with the specified features.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a front passenger seat upholstered in light brown leather with a diamond diamond-stitched pattern. which the description provided The seat angle across to reveal the of the seat, but the focus object,the passenger seat) and consistent with the description condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9b4f493eece31f84ecce8ce217273b495187e9ba
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7230e4b66a36332b1e959c6bcd68fa48a6ec3e90812b1bd86f356a8ccb6293ab
+size 155781
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..7f7b7f3b8d5edf2411dddb0327d05aacb07c5963
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with blonde hair and glasses, wearing a black jacket. She is engaged in a conversation with a man who is partially visible on the right side of the frame. The setting appears to be a television studio, as indicated by the presence of a logo in the top left corner and a microphone in the top right corner. The woman is speaking, and her facial expression suggests that she is making a point or sharing an opinion. The man is listening attentively, indicating that the conversation is serious or important. The overall style of the video is professional and polished, typical of a television interview or discussion."
+    ],
+    "video_ids": [
+        "Z0GcXkkgyAM_3_0to116"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with blonde hair and glasses wearing a black jacket, and a man partially visible on the right side of the frame.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman with blonde hair and glasses wearing a black jacket, and a man is partially visible on the right side of the frame. These elements match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a woman with blonde hair and glasses wearing a black jacket. and a man partially partially visible on the right side of the frame. The elements match the description provided, the questionObject(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d8998c3bcaea544a79d02230f95d53a6ad45dde4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:61bab1497f3674aa79c1927c3bb03f403e24bf8e758dd06ce03abc67c34c4b5c
+size 100577
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..47dbb411f34279cf6b370132c3f0a02a47fc19b8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a bald man with a beard and glasses, wearing a blue suit and a white shirt. He is seated in a room with a brick wall and a bookshelf in the background. The man appears to be speaking or reacting to something, as he is making a facial expression. The lighting in the room is soft and warm, creating a comfortable atmosphere. The man's glasses and beard give him a scholarly or intellectual appearance. The brick wall and bookshelf suggest a setting that could be a study or a personal library. The overall style of the video is realistic and naturalistic, with a focus on the man's facial expression and the room's ambiance."
+    ],
+    "video_ids": [
+        "7RJrnyB8jnQ_0_0to187"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bald man with a beard and glasses, wearing a blue suit and a white shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man with a beard and glasses, wearing a blue suit and a white shirt, which matches the description. The background and additional elements (like the Twitter handle overlay) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a bald man with a beard and glasses, wearing a blue suit and a white shirt. which matches the description provided The background appears additional elements dolike the book logo and) do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..54363bee236b68bb6cb666ebf205deba4ab1cbaf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1d8d590a72e1ab42bc84a1405a97243961b31457bd1f14c096f13ee6462adb5
+size 138248
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..3d1d43e4f7e849078e5acade9a35a1d3d4872665
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit, who appears to be in a serious or contemplative mood. He is looking off to the side, with a slight frown on his face. The background is blurred, but it seems to be an indoor setting with a patterned wall. The man is the main focus of the video, and there are no other significant objects or people in the frame. The style of the video is realistic, with a focus on the man's expression and the setting."
+    ],
+    "video_ids": [
+        "VeoiEqdkKAk_589_0to103"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit, looking off to the side with a slight frown.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a suit, and he is looking off to the side with a slight frown, matching the described condition. The background and other elements do not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit and which he is indeed off to the side with a slight frown. which the description condition. The attire is additional elements are not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ff91a92571fcfc2f9de5e20db998949571b3db4a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c65d1f644da6c60be652f072a3e755ff6e764b0d831a73f05e3fe716553ec96d
+size 104665
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..df6ba6c3c66d147c3c115edbf3399c7bd20d8b87
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated sequence featuring a group of superheroes. The style is realistic with a focus on action and character interaction. The scene takes place in an urban setting with a large building in the background. The superheroes are standing close together, suggesting a team dynamic. The characters are dressed in their respective costumes, which are detailed and colorful. The lighting is bright, highlighting the characters and their surroundings. The overall atmosphere is one of anticipation and readiness for action."
+    ],
+    "video_ids": [
+        "CRx1a2w4PZg_15_0to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A group of superheroes standing close together.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group of six superheroes \u2014 Thor, Hawkeye, Captain America, Iron Man, Hulk, and Black Widow \u2014 standing close together in a unified formation, which fulfills the 'Object(s)' condition. The background and additional elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a group of superheroes superheroes standing two, Captainkeye, Captain Marvel, Iron Man, and, and Black Widow \u2014 standing close together in a city pose. which alignfills the 'Object(s)' condition of The characters and additional characters, not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f73fb32d7a4c7760ced8d30e5f0937242c40d3c7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b1afefe45ef11e5af8c9bd34a88977a398b2cece53e9c81094743e8ee17b2eb
+size 193161
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..7cb452de2de1db796d03b7863eebf2f99e2be637
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a suit and tie standing in front of a transparent display case containing a large, intricately detailed model of an engine. The man appears to be explaining or demonstrating the engine, as he points to it with his right hand. The setting appears to be an indoor event or exhibition, as there are other people and exhibits in the background. The style of the video is informative and educational, likely aimed at showcasing the engine model and providing details about its design or function. The man's attire and the formal setting suggest a professional or academic context."
+    ],
+    "video_ids": [
+        "53ysOhWWFtw_11_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, a transparent display case containing a large, intricately detailed model of an engine.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a suit and tie standing in front of a transparent display case containing a large, intricately detailed model of an engine. The core elements described are present and accurately represented, with no significant contradictions. Additional background elements (such as other people and vehicles) do not conflict with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man in a suit and tie standing in front of a transparent display case. a large, intricately detailed model of an engine. The man elements of in present, match depicted in fulfilling no additional contradictions or The elements elements,like as other machinery and equipment) do not detr with the main description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f70e4673c501dd3377f94d2133a707e5234a61d7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1182-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3cafd6c2515275be22fc9f63143266f1ffb4987de5281070a73b255cf24bb86d
+size 123612
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..23f5f9309f234e7dcd52acc59e965ebcf0cf4dd2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a kitchen, waving his hand in a friendly manner. He is wearing a black t-shirt with the words \"Jesus Christ\" on it, and a white baseball cap. The kitchen is white, with white cabinets and a white refrigerator. The man is standing in front of a window, which is covered with blinds. The window lets in natural light, illuminating the room. The man appears to be in a good mood, as he is smiling and waving. The overall style of the video is casual and friendly."
+    ],
+    "video_ids": [
+        "RMDdhkkaDWU_4_26to162"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black t-shirt with 'Jesus Christ' on it, and a white baseball cap.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black t-shirt with 'Jesus Christ' printed on it and a white baseball cap, which matches the core description. The background and other elements do not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black t-shirt with theJesus Christ' written on it and a white baseball cap. which matches the description description. The setting and setting elements in not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a2157b2fe3984052f7781ddacebd957f33f6bd69
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0ec409ea20ab639dfe3335f86ce6b4f4cb57ff3c102ad271e2171f64da7b5079
+size 148382
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..904c863fd6daded1f04f296bad9116b56f5e8e8e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are standing on a grassy area near the ocean. The man is wearing a blue t-shirt and a cap, while the woman is wearing a brown tank top and has a flower in her hair. They are both holding a seashell, and it appears that they are examining it closely. The ocean is visible in the background, and there are rocks nearby. The overall style of the video is casual and relaxed, capturing a moment of leisure and exploration between the two individuals."
+    ],
+    "video_ids": [
+        "iXsVKqDXrZE_107_0to112"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman standing near the ocean. The man is wearing a light blue t-shirt and a cap, while the woman is wearing a dark tank top with a flower in her hair. They are interacting with each other, and the presence of additional elements like the ocean, rocks, and a grill does not contradict the core description of a man and a woman.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a man and a woman standing close each ocean, They man is wearing a blue blue t-shirt and a cap, while the woman is wearing a brown brown top and a flower accessory her hair. They are both with a other, with the presence of sheep elements like the ocean and sheep, and sheep flock in not contradict the core description of the man and a woman.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c7d997e88bf9f922f359282f8d703d2592b17598
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64622213b80e5a1d2c5b019053b1c406ba07f85a3791226a99b64b56a8ea8621
+size 115375
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..fe694e54a5b75fa022b6837140acfe77434c4d74
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows an elderly man sitting at a desk in an office setting. He is holding a black phone to his ear and appears to be in the middle of a conversation. The man is wearing a white shirt with a floral pattern and a black vest. The desk in front of him has a computer monitor and a red stapler on it. In the background, there is a window with yellow curtains and a pink wall. The overall style of the video is realistic and it seems to be a scene from a TV show or movie."
+    ],
+    "video_ids": [
+        "Z4xW6hzdDNU_22_0to181"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Elderly man, black phone, white shirt with floral pattern, black vest, computer monitor, red stapler? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an elderly man sitting at a desk, holding a black phone to his ear. He is wearing a white shirt with a floral pattern and a black vest. A computer monitor is visible on the desk in front of him, and a red stapler can be seen in the foreground. All specified objects are present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an elderly man wearing at a desk, holding a black phone to his ear, He is wearing a white shirt with a floral pattern and a black vest. There computer monitor is visible in the desk, front of him, and there red stapler is be seen on the foreground. The the objects are present and match the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..157da34a0556b86adc5cb8efbd8243da20a4dc7f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:684c16568d082d0ff50365df67b2485fbe68e9c0a395927526464322324c3804
+size 131780
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..dd8c496af673e45856b0f13664819f3a9c5df870
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a screenshot from a video game, specifically a football (soccer) match. The game is paused, and the user interface is visible. The scoreboard shows \"SL BENFICA 0 - 1 SPORTING CP\" with the time \"15 February 2020 16:00\" and the league \"Liga NOS\". The user interface includes options for \"Resume\", \"Instant Replay\", \"Trainer\", \"Match Facts\", \"Settings\", \"Forfeit Match\", and \"Exit\". The player's team formation is displayed on the left side of the screen, with player names and positions. The player's avatar is visible in the bottom left corner, wearing glasses and a yellow shirt. The background shows a football stadium with a green field and red stands."
+    ],
+    "video_ids": [
+        "BFgH7Dy7QlA_11_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Scoreboard, User Interface, Player's Team Formation, Player's Avatar? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly displaying a scoreboard (showing SL Benfica 0-1 Sporting CP), a user interface (with options like Resume Match, Instant Replay, etc.), a player's team formation (SL Benfica 3-4-3 Diamond layout with player names and positions), and a player's avatar (the man in the orange shirt in the bottom left corner). All these elements are present and clearly visible.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video contains fulfills the 'Object(s)' condition as displaying showing a scoreboard,showing the vsfica 0-1 A CP), a user interface withwith various like ',, Replay Replay, etc.), a player's team formation (visible Benfica's4-4-3,),), player positions and positions), and a player's avatar (a character in the yellow shirt). the bottom right corner). The these elements are present and match visible in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d76b26c7d0e9a155ab8f0520ede173019c38bb22
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95ed68e188f9d79c59efd6346744e0fb069b01679e18933b0cc66fcc4a62e431
+size 165916
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..694bd0915cbf43a1773b6d7ceac260c8a83fb043
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the process of pouring a rich, brown curry from a metal pot into a bowl. The curry is thick and appears to be a type of Indian dish, possibly Dal makhani, as indicated by the text overlay. The pot is placed on a metal tray, and the curry is being poured into a bowl that is positioned to the right of the pot. The background is blurred, but it seems to be a kitchen setting with other pots and pans visible. The style of the video is simple and straightforward, focusing on the action of pouring the curry without any additional context or embellishments. The lighting is bright, highlighting the texture and color of the curry."
+    ],
+    "video_ids": [
+        "VfNqA2ukNrk_182_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A metal pot containing curry and a bowl positioned to its right.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a metal pot containing dal makhani curry, and to its right, there is a bowl with sliced onions, which matches the described objects. The presence of other elements like additional bowls and utensils does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a metal pot containing a (akhani, being which a the right, there is a bowl positioned the green and which is the description setup. The curry of additional elements like a pots and ails does not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4fdc08a1ed3d9dfbe3e21ee6ae72d6d164be44f5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c0f9ca359b47f65d30c2c2c343aa6dcca602eb14e9c24f0b15dc60ecbb700e6c
+size 214901
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..d929fec545f68e98497be5c9178086eabaf0725c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a group of people in a forest setting. The central figure is a woman wearing a pink shirt and a red headscarf. She is sitting and appears to be engaged in a conversation with the others. The group consists of both men and women, all dressed in casual clothing. The forest around them is lush and green, with trees and foliage filling the background. The lighting suggests it is daytime. The style of the video is naturalistic, capturing a candid moment among the group in their natural environment."
+    ],
+    "video_ids": [
+        "JP6WMSvFS20_194_146to275"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a pink shirt and red headscarf, a group consisting of both men and women dressed in casual clothing? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a pink shirt and a red headscarf, sitting in a group that includes both men and women dressed in casual clothing, as described. The background and other individuals are consistent with the scene, and no elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group wearing a pink shirt and a red headscarf, and among a group of includes both men and women dressed in casual clothing. which described. The setting is setting elements in consistent with the casual being and there additional contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8b49f0377abb1136ba651f7b98b2e5a53f0fc707
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e20ba19aaac09d4655a1c40b89cabeb3e9b62710e5b7148ac46aaa587f86205f
+size 166999
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..8dd142166d7c257113339d603daa033fa29a0494
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a television show segment featuring a man in a blue and white hoodie. He is holding a book and appears to be speaking or presenting. The background is a blue wall with a truck graphic. There are subtitles in Korean, indicating that the show is likely in Korean. The style of the video is a standard television show format with a focus on the man and his book. The show seems to be informative or educational, as suggested by the man's attire and the book he is holding. The overall tone of the video is casual and informative."
+    ],
+    "video_ids": [
+        "aKWenpfemlE_124_169to293"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a blue and white hoodie holding a book? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue and white hoodie, holding what appears to be a book or a similar rectangular object. The description is accurate in terms of the core elements: the man's attire and the object he is holding. The presence of text overlays and background elements does not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue and white hoodie, holding an appears to be a book. a stack object object. The description matches largely in terms of the man elements: the man's attire and the object he is holding. The background of a and and a elements does not contradict the core.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c76c7b1a8e34c21f1ffac29aac90ea75ef5ef627
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b9c94433ea7973fe94003ee77be1e24ceecb6c3de152a56d697e33b02b856c5
+size 165744
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..c3cf5447f73984591c1d83b3375e40acb8ae4bc8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car, focusing on the back seat. The car has a luxurious design with brown leather seats. The back seat is empty, and the seat belts are visible, indicating that the car is ready for passengers. The car's interior is well-lit, and the seats are neatly arranged. The style of the video is straightforward, with no additional elements or distractions. The focus is solely on the car's interior, providing a clear view of the back seat and its features."
+    ],
+    "video_ids": [
+        "PD9cQYXzRWY_84_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Back seat, seat belts? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the back seat of a car with visible seat belts, matching the specified object condition. The focus remains on the back seat and its seat belts throughout the frames, with no conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows the back seat of a car with leather seat belts. which the description ' condition. The image is on the back seat and seat features belts, the frames, indicating no additional elements present}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8b24798cb607471f95b2bb4fd2090c91af7b3a22
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1183-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83220c205e4dfc08c9bc8a8445bfe1ce18511b993731386e5760994f391d4ad5
+size 132493
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..767c1d5c4b7f3a96d9879210062689f37887ae7e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the beauty of nature with a close-up view of three vibrant red flowers in a field of green grass. The flowers, with their delicate petals and intricate patterns, are the main focus of the video. The first frame shows the flowers in full bloom, their petals fully unfurled and their colors at their brightest. In the second frame, the flowers begin to close up, their petals starting to fold in on themselves. By the third frame, the flowers have fully closed up, their petals tightly packed together, creating a contrast between the open and closed states of the flowers. The video is a beautiful representation of the cycle of life and the beauty of nature."
+    ],
+    "video_ids": [
+        "gn9_R1eeFac_6_0to174"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three vibrant red flowers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features three vibrant red flowers in focus, which matches the 'Object(s)' condition. While there are additional flowers and greenery in the background, they do not contradict the core description and are consistent with a natural garden setting.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features three vibrant red flowers, the, which align the 'Object(s)' condition. The there are additional elements and greenery in the background, the do not detr the core description of serve secondary with the natural setting setting.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7cbb4d5c24ad2aa803aa6648a216597ce92d3471
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:afcbbe0cb39d516d0717c811e4ef4982950b8499d3fd35dc099ead7cdb5b6da7
+size 124573
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..e6cbc0ea46a2bacc6b0bfb05f977ab2d0339734f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man standing by a lake, wearing a white cap and sunglasses. He is looking out at the water, which is calm and blue. The man appears to be in a relaxed state, possibly enjoying the view or waiting for someone. The surrounding environment is peaceful and serene, with trees and bushes lining the shore. The man's attire suggests that it might be a sunny day, and he is prepared for outdoor activities. The overall style of the video is casual and natural, capturing a moment of tranquility in a beautiful outdoor setting."
+    ],
+    "video_ids": [
+        "Tl1uM5viU2U_8_231to401"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man standing by the lake, wearing a white cap and sunglasses.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man standing by a lake, wearing a white cap with sunglasses resting on top of it. The core description is accurately represented, with no significant contradictions. Additional elements like trees, water ripples, and a logo in the corner do not conflict with the primary subject description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man standing by a lake, wearing a white cap and a, on top of it. The setting elements of accurately represented in with the significant contradictions or The elements like the and water,ipples, and a clear on the background do not conflict with the main description and.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..36154849869cedac9300308600e472ecd5bb96d6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b42bb83884f8e643fca2408e8d26d908fa4c51b3452021d16f59239a528d3593
+size 121859
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..b4864f53002580613ac94a3bcc9f53c4ddc8ee1e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in a modern office space. He is dressed in a dark suit and a white shirt, with a beard and mustache. The office has large windows that let in natural light, and there are green and blue ottomans in the background. The man appears to be speaking or presenting, as suggested by the presence of a microphone. The overall style of the video is professional and polished, with a focus on the man and his surroundings."
+    ],
+    "video_ids": [
+        "eNccFQLOzlI_1_109to256"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man dressed in a dark suit and white shirt, with a beard and mustache, holding a microphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The man in the video is wearing a dark suit and white shirt, has a beard and mustache, and is speaking with a microphone clipped to his suit. These elements match the description provided. The presence of other objects like bean bags and a lamp in the background does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video in the video is dressed a dark suit and a shirt, which a beard and mustache, and is standing, a microphone in to his suit. The elements match the description provided, The setting of the elements like the bags and a window in the background does not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6cb3baf4c72ca03c4084abcf3bf43787a913a511
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c66a7864f141a9e31674f6c99d406279951ad733061c842997d4d8b801e62dfb
+size 154588
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..92b44713a754c8e096f2d6b5f9d9afd7c4feeca6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the iconic Statue of Liberty in New York City. The statue, a symbol of freedom and democracy, stands tall against a cloudy sky. The statue's green patina, a result of oxidation, gives it a weathered and timeless appearance. The golden flame at the top of the statue's torch is a striking contrast to the green of the statue. The video is taken from a low angle, emphasizing the statue's height and grandeur. The perspective also allows for a view of the statue's base, which is not visible from ground level. The overall style of the video is a realistic portrayal of this world-renowned landmark."
+    ],
+    "video_ids": [
+        "Ei8SAMhNjWU_22_0to155"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Statue of Liberty, cloudy sky? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features the Statue of Liberty, showing its torch and crown in detail, and the background is a cloudy sky, matching the specified conditions. The presence of a year overlay (1902) and a watermark (SONNEN SEITE) does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features the Statue of Liberty, which it iconic and crown, a, which the sky is filled cloudy sky, which the description conditions. There presence of the building-round does2980) does the person doesCYENLANDHEN) does not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..55fa7e0ac29d95fe31b8213b23ff27227584e57b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd80d5ea8ffdaf1f862c6c86a6da0fdab3ecc4f42cf12e4f2d1617c0916921ad
+size 75406
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..e0b95c2bdf2ddda6983d3938c00db48c31b41b78
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men in a kitchen, with one man wearing a blue and white checkered shirt and an apron, and the other man wearing a black shirt. They are standing in front of a refrigerator and a wooden cabinet. The man in the apron is looking down, possibly at a recipe or a cooking task, while the other man is looking at him. The kitchen appears to be well-lit and clean, with a modern design. The style of the video is realistic and seems to be a scene from a cooking show or a home cooking video."
+    ],
+    "video_ids": [
+        "L8jmeDqtNjM_3_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, a refrigerator, a wooden cabinet, a blue and white checkered shirt, an apron, a black shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men, one wearing a blue and white checkered shirt with an apron and the other partially visible. A refrigerator and a wooden cabinet are visible in the background. The description includes a black shirt, which is not clearly visible as part of the main subjects' attire, but since the core elements (two men, refrigerator, wooden cabinet, checkered shirt, apron) are present and match the description, the video largely fulfills the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts two men in one wearing a blue and white checkered shirt and an apron and the other in visible in There refrigerator and wooden wooden cabinet are present in the background. The man accurately a blue shirt, which is worn directly visible in the of the attire subject but attire but but the the core elements matchtwo men, refrigerator, wooden cabinet, blueered shirt, apron) are present, match the description, the answer largely fulfills the '.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..63f299a031b76d7448d47d6af55002e5ae07b272
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1bad5051c92e2765109b4a07d35f367c53d9c0621ce4e752be2ae0d902c3dca6
+size 360504
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..a481d6b5febf71e4ffe5904a53eea328930a9500
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a kitchen, standing in front of a refrigerator. She is wearing a blue dress and a necklace. The kitchen is well-equipped with various utensils and appliances, including a sink, a stove, and a refrigerator. The woman appears to be in the middle of cooking or preparing a meal. The overall style of the video is casual and homey, capturing a typical day in a kitchen."
+    ],
+    "video_ids": [
+        "0AhUNNSKZvI_2_0to113"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a blue dress with a necklace, standing in front of a refrigerator.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a blue dress with a necklace, standing in front of a refrigerator. The background includes kitchen elements like a countertop, eggs, and a fruit bowl, but these do not contradict the core description. The woman's attire and position relative to the refrigerator match the specified condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a blue dress and a necklace, standing in front of a refrigerator. The setting includes a elements such a stoveop with a, and a bowl bowl, which these are not contradict the core description. The woman is attire and the in to the refrigerator match the description conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2d7d05a5cad9ce79b17d15c6a372f8c72954b9dd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c21da83f65da063af83f0b0b8c2e28765231459153223459f4d220621144a1f
+size 204456
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..91cba7561ad7b23feb4adc4768e7b2723f5c6441
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a group of people walking along a wooden boardwalk that hugs the edge of a cliff. The boardwalk is situated above a rocky shoreline, where the waves crash against the rocks, creating a dynamic and powerful scene. The sky is overcast, casting a soft light over the scene. The people are dressed in casual clothing, some carrying backpacks, suggesting they are out for a leisurely walk or hike. The boardwalk is made of wooden planks, and it appears to be sturdy and well-maintained. The overall style of the video is naturalistic, capturing the beauty of the landscape and the casual, unhurried pace of the people walking along the boardwalk."
+    ],
+    "video_ids": [
+        "SokSuP4Y9Zw_31_84to282"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A group of people, wooden boardwalk, rocky shoreline, waves, sky? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a group of people walking along a wooden boardwalk, a rocky shoreline with waves crashing against it, and an overcast sky. All specified elements are present and accurately depicted without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a group of people walking on a wooden boardwalk. which rocky shoreline, waves crashing against it, and a expansivecast sky. The these elements are present and contribute depicted, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fb939d4b547c4b7452edc8d11f45944bec64bd9f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f0236f99e144ff31f4261e695d19d3bd08eb07fbe772043030c7500b9fb3f258
+size 169757
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..009cdaf55593195661abdee6bbce19fab24ac0c5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a group of construction workers in orange and yellow safety gear, working on a street. The workers are seen in various positions, some standing, some kneeling, and one even sitting in a hole. They are using a large yellow machine, which is likely a part of their construction equipment. The scene is set on a street, with a truck parked nearby. The workers are focused on their task, indicating that they are in the middle of a project. The video provides a glimpse into the daily work of these construction workers, showcasing their teamwork and dedication to their job."
+    ],
+    "video_ids": [
+        "FJ5k--ucOuU_30_0to110"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Construction workers in orange and yellow safety gear, a large yellow machine (likely construction equipment).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple construction workers wearing orange and yellow safety gear, including hard hats, and a large yellow machine (likely a pipe-laying or trenching machine) positioned near a manhole. These elements match the described condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts construction construction workers wearing orange and yellow safety gear, actively reflective hats and reflective they large yellow machine,likely construction construction-laying machine roading machine) in in the grouphole. The elements align the description ',}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6e7a25376f79f175363be6b0a2dd43d4cbfe7534
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1184-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:270d15668d2cc4a80f0b528d36f6d2f1056ae288460df9975ce32e57cdc70d81
+size 172026
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..f62d652016137c9d1d5284203d639e865043f274
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a long beard and a casual demeanor, standing in a forested area. He is wearing a gray t-shirt and appears to be gesturing with his hands as he speaks. The forest around him is lush and dense, with trees and foliage filling the background. The lighting suggests it might be late afternoon or early evening, with the sun casting a warm glow on the scene. The man's relaxed posture and the natural setting give the video a peaceful and serene atmosphere."
+    ],
+    "video_ids": [
+        "O66jIgc-dtI_28_0to172"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a long beard and a casual demeanor, wearing a gray t-shirt, gesturing with his hands as he speaks.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a long beard, wearing a gray t-shirt, standing in a forested area with a tent in the background. He is gesturing with his hands while speaking, which aligns with the described 'Object(s)' condition. The setting and his demeanor are casual, and no elements contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a long beard and wearing a gray t-shirt, and in a naturaled area. lush lush in the background. He is gesturing with his hands as speaking, which aligns with the description 'cas(s)' condition. The setting and the actions match consistent, matching the additional contradict the description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..53249c21d74f9b12fcb6d78b378cc8b5123678ec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:484f20423155e8cf43a624afef9233499eaf52a122c4d731dae8658ca50ead08
+size 190152
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..2a2065c3fdc1331c0688b91ffc9ba4afa00f6e9c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen enjoying a large portion of French fries in a red and white checkered paper container. He is seated in a car, with the window open, allowing a glimpse of the outside world. The man is dressed casually, and his focus is solely on the delicious fries he is about to enjoy. The video captures the simple pleasure of indulging in a favorite snack while on the go."
+    ],
+    "video_ids": [
+        "NRf6lZYzxU0_37_36to174"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a large portion of French fries, a red and white checkered paper container? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man sitting in a car, holding a red and white checkered paper container filled with a large portion of French fries. The core elements described are all present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man holding in a car, holding a large and white checkered paper container filled with a large portion of French fries. The man elements of in present present and accurately depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..85f4d1a42f842a558f33c4eb2b876e8d2d4847fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0a922868740b9e68f653f0283a6a5af1d3383a36d3f1231a1ed4fe0e5661a39
+size 182937
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..1c3fdee758abbcd9e6d18528326b87d7b8af0457
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is an aerial view of a town in Chile, specifically Puerto Varas. The town is situated on the shore of a large body of water, possibly a lake or a bay, with a mountain range visible in the distance. The town is densely populated with buildings of various sizes and colors, including red, yellow, and white. The streets are lined with trees and grassy areas, and there are cars parked along the sides of the roads. The overall style of the video is a realistic, color-rich depiction of a small town in a beautiful natural setting. The video captures the town's layout and the surrounding landscape, providing a comprehensive view of the area."
+    ],
+    "video_ids": [
+        "9E2TjdcLep8_7_0to153"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Buildings, streets, trees, grassy areas, cars? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing buildings, streets, trees, grassy areas, and cars in the aerial view of Puerto Varas, Chile. These elements are prominently featured and accurately represent the described objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing buildings, streets, trees, grassy areas, and cars. a scene view of a Nas. Chile. The elements are prominently featured and align represent the described scene.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..98c3afe8792ed3ba0b0d1ba03b6616e340207c7f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5f0907f5298c144373e0ab58116b06cad80a1ae1ff6b9bf62d65e17dd4e0f23
+size 240115
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..598e27ec74286e1789fe127aeffe8e584d90b3b5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features three women standing in a row, each with a unique hairstyle and outfit. The woman on the left has short brown hair and is wearing a purple dress with a black collar. The woman in the middle has long blonde hair and is wearing a white dress with a blue collar and a necklace. The woman on the right has short brown hair and is wearing a yellow shirt with a black collar and a star pattern. They are all smiling and appear to be engaged in a conversation. The background shows a blue sky and a building with a brown roof. The style of the video is casual and friendly, with a focus on the women and their interaction."
+    ],
+    "video_ids": [
+        "JHFbxVb7s8Q_4_0to174"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three women standing in a row? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows three women standing in a row, facing the camera. Their positioning and alignment match the description, and no elements contradict this core observation. Additional background elements (buildings, trees) are present but do not interfere with the primary subject.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows three women standing in a row. which the camera. They attire and the fulfill the description of and there additional contradict this core condition.\"\n The elements elements,likeings and trees) do present but do not interfere with the main focus of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0219cb10e14d40d24b9d13e7c5e0a207c29b681c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dde162438077ec80d08e296839242e3e33033497203b4ab1820f0f43ee73c5ff
+size 136869
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..5a9d51c6df71d29701bfa03172495eaf8c0870e2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a culinary tutorial featuring a pink smoothie. The style is clean and modern, with a focus on the vibrant color of the drink. The presenter, Sandra Matarazzo, is not visible in the video. The production and editing are done by Receitas e Tempos. The video shows a close-up of the pink smoothie in a clear glass, with a white straw and a white spoon. The smoothie is placed on a wooden table, and there are slices of pink grapefruit and green limes in the background. The overall atmosphere of the video is fresh and appetizing, inviting viewers to try the recipe."
+    ],
+    "video_ids": [
+        "GRCK_NcjFmw_21_16to160"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Pink smoothie, clear glass, white straw, white spoon, pink grapefruit slices, green limes? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a pink smoothie in a clear glass, with a white straw inserted. Two white spoons are visible on the wooden surface, and there are pink grapefruit slices (one cut in half) and green limes in the background. All specified elements are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showcasing showing a pink smoothie in a clear glass, a a white straw and. There slices spoons are visible, the table surface, and there are pink grapefruit slices andone whole in half) and green limes in the background. The these objects are present, match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c95eb8c5bc02cde08f3b9fc451b0551ad6ef11da
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0b334b6c49c23bc08a169fda5bcad3d998ffce07d4a7ea8f7b2a0766d69c31f
+size 45392
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..b9b89642dbdb365759a7350ca9a75b3eb53a8d1e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a white unicorn and a lion with brown fur are seated at a dining table. The table is set with plates, cups, and various food items such as fruits and cakes. The unicorn and lion appear to be enjoying a meal together. The setting is a room with a yellow floor and a white wall. The overall style of the video is cartoonish and whimsical, with a focus on the two main characters and their interaction with the food on the table."
+    ],
+    "video_ids": [
+        "M-3Elxy46tw_188_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white unicorn and a lion with brown fur.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a white unicorn with a golden horn and a lion with brown fur sitting at a table. These two main characters are prominently featured and match the description exactly. Additional elements in the background (other cartoon characters) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a white unicorn and a golden horn and a lion with brown fur sitting at a table. The are animals characters match the featured, match the description provided. The elements like the video,like chairs animals, do not conflict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..af77c2465a7b2cbdb2bfab4af650d811d338e425
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb1c59e3eae49eccce3daa95a93849a5d4a5e7b2b3e9f257b96b0121b4f1c8ad
+size 130912
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..a8f016c7dc7fd0592313712dfd975813be2a6126
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a car decorated with festive Christmas reindeer antlers and red nose, parked in a driveway. The car is a dark color, possibly black or dark grey. The reindeer antlers are large and prominent, with the red nose being a bright, eye-catching color. The car is parked in front of a house with a garage, and there are palm trees visible in the background, suggesting a warm climate. The video is likely taken during the holiday season, as indicated by the car's decorations. The style of the video is casual and informal, with a focus on the car and its decorations. The camera angle is from a low perspective, looking up at the car, which adds a sense of scale and emphasizes the size of the reindeer antlers. The video does not contain any text or additional graphics."
+    ],
+    "video_ids": [
+        "Fh6UoxFta20_36_0to157"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A car decorated with festive Christmas reindeer antlers and a red nose? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a car decorated with festive Christmas reindeer antlers on the roof and a red nose on the front grille, matching the description. Additional decorative elements like bows and lights do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a car decorated with festive Christmas reindeer antlers and the hood and a red nose on the front of. which the description provided The elements elements like the and a are not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..162094745ce4208a27a37cea8fd958b382dc1cf7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b906674dd79f4869051d9a0e964320421f1ca0bb7eb3bc8b03516341d218056
+size 169229
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..cdf84b16bc8c57ecbb907bd892df6d372977fbc1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video begins with a close-up shot of a table set for a meal. On the table, there is a plate with roasted chicken garnished with broccoli and cherry tomatoes, a glass of red wine, a bottle of red wine, and a lit red candle in a small holder. The scene is calm and inviting, suggesting a quiet evening or celebration.  Suddenly, the bottle of wine is struck by an unseen force, causing it to shatter dramatically. The explosion sends shards of glass and wine splashing everywhere, creating a chaotic burst of motion. The camera remains stationary throughout, capturing the full extent of the explosion from a fixed position. The initial stillness of the table is abruptly disrupted by the violent impact, transforming the serene setting into one of sudden and intense action."
+    ],
+    "video_ids": [
+        "c55c04e849d40e4da92755acbaa832087b5a2f18733eda26b89197fe9e5ec044"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A plate with roasted chicken, broccoli, and cherry tomatoes; a glass of red wine; a bottle of red wine; a lit red candle in a small holder.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a plate with roasted chicken, broccoli, and cherry tomatoes; a glass of red wine; a bottle of red wine; and a lit red candle in a small holder. These objects are clearly visible in the initial frame and are present before the explosion, fulfilling the 'Object(s)' condition. The explosion does not contradict the presence of these items, as they are shown intact before the destruction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows ful a plate with broccoli chicken, broccoli, and cherry tomatoes, a glass of red wine; a bottle of red wine; and a lit red candle in a small holder. The elements are all visible in the video frames, remain consistent in the dramatic effect which the 'Object(s)' condition.\"\n The explosion does not contradict the core of these objects, as it are not before before the event.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..076c4fdc84c9d122cdc154f228ed3cd5480e5766
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1185-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:702a26f70ee504e3642e1af51896767f1c6495154d4dfc93d8a547d71109bdb9
+size 254582
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..684871fbba4ba40e1d7e0e8ee92e29735ff8e3f6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a close-up of a car window with a decal. The decal is a shield-shaped logo with the text \"Russell Built\" at the top and \"Custom Motorsports Fabrication\" at the bottom. The logo is predominantly black with red and white accents. The style of the video is a straightforward, static shot with no movement or action. The focus is on the decal, which is the main subject of the video. The background is blurred, emphasizing the decal on the car window. The video does not contain any additional objects or elements. The overall style of the video is simple and straightforward, focusing on the decal and its details."
+    ],
+    "video_ids": [
+        "2EwNUsO9dAM_10_0to133"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: car window decal? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a decal on a car window, which matches the 'Object(s)' condition. The decal features the text 'RUSSELL BUILT Custom MOTORSPORTS FABRICATION' and is prominently displayed on the window. There are no conflicting elements that contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a caral on a car window that which matches the descriptionObject(s)' condition. The decal is the text 'RussessSELL BRY' MotorcycleCYS'ABRICICATION' and is positioned displayed on the car, The are no additional elements in would the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5ce26815e0dc1eee2a287b8e214af65beffdaff5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f07361a77b179e1a16b8266e861ca9cec5cf478c2e27c61e1da1460c22032f92
+size 120116
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..b1ecfe9f5db2b68c8dec250227c7d0433eba23cf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a black Bentley car driving on a race track. The car is sleek and shiny, with a distinctive grill and round headlights. The track is surrounded by a fence and has a few grandstands in the background. The car is moving at a high speed, as indicated by the blurred background. The video is likely a promotional or review video for the Bentley car, showcasing its performance and design. The style of the video is dynamic and action-packed, with a focus on the car's speed and the excitement of racing."
+    ],
+    "video_ids": [
+        "Rf_Hk47k-r0_26_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black Bentley car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a black Bentley car driving on a racetrack, which matches the core description. The car's design, color, and brand are clearly visible and consistent with the requested object. Additional elements like the track, fencing, and background structures do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a black Bentley car, on a tracketrack. which align the description description of The car's design, including, and the are clearly visible, consistent with the description object. The elements like the rac and barriers, and trees do do not contradict the main but}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7360727bbc6daba37a4287b83835f5a828cce021
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:61087001dcdffb60ebc10e2dfa0484332dd756374111811da99dcb163a3780c9
+size 266453
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..d56f772140aaa605225e537e2ce661509d71a569
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person's hand gently holds a small turtle with a grayish-brown shell and a white head. The turtle is positioned in the center of the frame, drawing attention to its unique appearance. The background features a lush green plant with vibrant leaves, providing a natural and serene setting. The overall style of the video is calm and peaceful, capturing a moment of interaction between the person and the turtle in a tranquil outdoor environment."
+    ],
+    "video_ids": [
+        "8pjzJYZChxI_13_0to178"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person's hand holding a small turtle with a grayish-brown shell and a white head.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person's hand holding a small turtle. The turtle has a grayish-brown shell with visible scutes and a white head with speckled patterns, matching the description. The background includes green foliage, which does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person's hand holding a small turtle with The turtle has a grayish-brown shell and a patternsutes and a white head. ackles patterns. which the description. The background includes green leaves, which is not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..74ca7b11c099fc90052d311eeed2b6bb8cf348df
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6f9f8e23cdde892b313ccdde802b32240376fb56e0014262f73b186569bac6c
+size 108995
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..352a1b0f285c03a044fa39f59f54d22b0832f806
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a plate of spaghetti with toppings, presented in a realistic and appetizing style. The spaghetti is yellow and appears to be cooked al dente, with a glossy sheen that suggests it has been coated with a sauce. The toppings include chopped nuts, which are scattered across the spaghetti, adding a contrasting texture and color. There are also green herbs sprinkled on top, providing a fresh and aromatic element to the dish. The plate is white, which contrasts with the colors of the spaghetti and toppings, making the dish stand out. The background is blurred, drawing focus to the plate of spaghetti, and the lighting is soft and warm, enhancing the visual appeal of the dish. The video does not contain any text or additional elements, and the focus is solely on the plate of spaghetti. The style of the video is realistic and detailed, with a focus on the textures and colors of the food, making it an effective advertisement for the dish."
+    ],
+    "video_ids": [
+        "sVFXO7moET0_0_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Spaghetti, chopped nuts, green herbs? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a plate of spaghetti topped with chopped nuts and green herbs, matching the described objects. The focus remains on these elements throughout the video, with no contradictory elements present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a plate of spaghetti with with chopped nuts and green herbs, which the description objects. The presence is on the elements, the frames, with no additional elements present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e25264c576006e3a5f26d5d208dea57db7c85e41
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b8d7a3d21b094aca07967cac47d30ad493594a88391805bf857e80b0cb8944a8
+size 59750
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..a0c8d332ae59fcbb9df38a6e2d5962264fa3f2a5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman sitting in a chair with a microphone in front of her, wearing headphones and a black jacket with white stripes. She is smiling and appears to be engaged in a conversation or recording a podcast. The room has a red wall and a couch in the background, and there is a dog lying on the couch. The style of the video is casual and informal, suggesting a relaxed and friendly atmosphere."
+    ],
+    "video_ids": [
+        "MKzl8E_AGlM_2_20to228"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman sitting in a chair with a microphone, wearing headphones and a black jacket with white stripes, and a dog lying on the couch.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman sitting in a chair with a microphone and headphones, wearing a black jacket with white stripes, which matches the description. A dog is also visible lying on a couch in the background, fulfilling that part of the condition. There are no contradictions with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman sitting in a chair with a microphone, wearing, wearing a black jacket with white stripes. which matches the description. Additionally dog is also visible lying on the couch in the background, which the part of the description. The are no contradictions with the given description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..09db0ef604eb5124559739c7e4212e74d12e0e31
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d65205d71f44c6dd5c00fde2d4112fecca6a4d8834e9ff28abda29aafdd7408b
+size 181015
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..123f4880a00107f3c7a16e325e2d3a88ff6a690a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a vibrant and colorful display of a religious shrine. The central focus is a large, ornate statue of a deity, adorned with a rainbow halo and surrounded by an array of flowers. The flowers, in hues of red, yellow, and white, are arranged in a semi-circle around the statue, creating a sense of reverence and devotion. The shrine is set against a backdrop of a blue sky, adding to the serene and peaceful atmosphere. The overall style of the video is one of celebration and reverence, capturing the essence of the religious experience."
+    ],
+    "video_ids": [
+        "78T5pogEEAU_41_0to116"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large, ornate statue of a deity with a rainbow halo, surrounded by flowers.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large, ornate statue of a deity (appearing to be a Buddha or similar figure) with a colorful, rainbow-like halo behind it. The statue is surrounded by various flowers, including red and yellow blooms, which matches the description. The presence of additional elements like framed pictures and candle holders does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a large, ornate statue of a deity withlikelyaring to be a Buddhist or a figure) with a rainbow, rainbow-like halo. it. The statue is seated by a flowers, which red, yellow blooms, which align the description provided The overall of the elements like the flowers or a-like does not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9c96160d65994e45c886389d4aef02018a10ac2c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5ac57dffdaa4ac4c9cf1145b1d550911498b4f129e79dd605d05c3ae1dfd4412
+size 165124
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..260a03b85d8d9e7a5c5abd6daa169c6809376c85
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene moment on a mountain top where two individuals are standing near a stone pagoda. The pagoda, with its multiple tiers, stands as a focal point in the scene. The individuals, dressed in casual attire, are positioned on a stone platform that is surrounded by a white railing. The platform offers a panoramic view of the surrounding landscape, which is lush with green trees and rolling hills. The sky above is a clear blue, suggesting a bright and sunny day. The overall style of the video is naturalistic, with a focus on the beauty of the landscape and the peaceful interaction between the individuals and their environment."
+    ],
+    "video_ids": [
+        "35EHFIIrw5c_20_21to162"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two individuals and a stone pagoda.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two individuals standing near a stone pagoda on a mountainous landscape, fulfilling the core description. The presence of additional elements like the railing and background scenery does not contradict the main condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts two individuals standing on a stone pagoda. a platformous landscape. which the ' description of The presence of additional elements like the mountain and the scenery does not contradict the main condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c5faf2d0e845f24f90f41f5cef01f7446d6ecbea
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b757cb2014207f19298772c427894e7269856e3f8715afe115d0735e1b3a36e
+size 123754
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..d60aff87ccc6c912853d9343aea356bcb0bd7a74
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen interacting with a modern coffee machine. The machine is sleek and silver, with a digital display and multiple buttons. The person, dressed in a blue shirt, is pointing at the machine, possibly selecting a coffee option or adjusting settings. A cup is placed on the machine, ready to receive the coffee. The scene suggests a casual, everyday moment, perhaps in an office or a coffee shop. The focus is on the interaction between the person and the coffee machine, highlighting the convenience and ease of use of modern coffee-making technology."
+    ],
+    "video_ids": [
+        "q18tLLSFZuQ_5_0to182"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person in a blue shirt and a coffee machine.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person wearing a blue shirt interacting with a coffee machine. The person is seen placing a cup under the machine's dispenser and pressing buttons on the machine's control panel. The coffee machine is prominently featured and appears to be a commercial-grade model with a digital display and multiple beverage options. There are no elements in the video that contradict the description of a person in a blue shirt and a coffee machine.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a person in a blue shirt interacting with a coffee machine. The person is seen placing a cup under the machine, sp, appears a, the machine's digital panel, The presence machine is prominently displayed, appears to be the modern-grade model, a digital display. control buttons options. The are no additional in the video that contradict the description of a person in a blue shirt and a coffee machine.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a1cee037c0d0cce55ee441a350f11d41a56cd819
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1186-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:387f19d0161b0fe4352a36d829b9ec6d8444e79d276dde066e0ed7e5a03a3670
+size 162200
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..f18c021d046bf431b74f2c3eb528c927b2cb9d66
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a promotional advertisement for the Simpson Ghost Bandit Helmet, which is available now at Revzilla.com/april16. The helmet is prominently displayed in the center of the image, with a close-up view of its black and white design. The helmet is placed on a table, and a pair of black gloves is placed next to it. The gloves are also black and white, matching the helmet's design. The background of the image is blurred, but it appears to be a workshop or garage setting. The overall style of the video is sleek and modern, with a focus on the helmet's design and features. The video is likely aimed at motorcycle enthusiasts who are interested in purchasing the helmet."
+    ],
+    "video_ids": [
+        "MjYSursspZ8_34_0to220"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Simpson Ghost Bandit Helmet, black gloves? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features the Simpson Ghost Bandit Helmet and black gloves as central objects on a table. The text overlay also confirms the helmet's identity. Additional items like a coiled cable and a box are present but do not contradict the core description, so the video successfully fulfills the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a Simpson Ghost Bandit Helmet, black gloves, described elements. a table. The helmet ' on matches the presence's model. The elements like a blurrediled cable and a small in present but do not conflict the core description of as the video largely fulfills the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..db5a3de9ec8b00d5a75b0dc3d7b2946217e22437
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6a59984bc4ecb247389adf54427b24f7d93e49e76fd9f3a5291ca18af30f8a8
+size 94595
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..a4a01179bd8c6866c484bdfab0143be01ab1a271
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the iconic Sydney Opera House, a renowned landmark in Australia. The first frame shows the Opera House from a distance, with the city skyline in the background. The second frame provides a closer view of the Opera House, highlighting its distinctive white sail-like roofs. The third frame focuses on the water surrounding the Opera House, with boats gently floating on the surface. The video style is a combination of aerial and ground-level perspectives, providing a comprehensive view of the Opera House and its surroundings. The colors in the video are vibrant, with the white of the Opera House contrasting against the blue of the water and the green of the city. The overall mood of the video is serene and peaceful, capturing the beauty of this world-famous landmark."
+    ],
+    "video_ids": [
+        "wGDb4NACTfM_10_57to187"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Sydney Opera House, city skyline, boats? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by prominently featuring the Sydney Opera House, a clear city skyline in the background, and multiple boats (ferry and smaller vessel) moving through the water. All specified elements are accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting featuring the Sydney Opera House, which city city skyline in the background, and boats boats onaerry and smaller vessels) on across the water. The elements elements are present represented, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..743fc82be7cf64651a6a27bd4fc52fd0595b538a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef8e4a0cc5eeb97811eee80856e64f3d95b976433d018b11a58ddd174bc0a9ff
+size 196586
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..6d12398ba08c0845be26bbe1ec25719585055828
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a futuristic cityscape with a large, black, cylindrical object that appears to be a spacecraft or a large pipe. The object is situated in the middle of the city, surrounded by tall skyscrapers. A set of stairs leads up to the object, suggesting that it is accessible. The city is bustling with activity, and the sky is clear, indicating a sunny day. The overall style of the video is realistic with a touch of science fiction, as evidenced by the presence of the spacecraft-like object. The video likely explores themes of urban life, technology, and the future."
+    ],
+    "video_ids": [
+        "QbsOXfIXFfk_1_73to245"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large, black, cylindrical object resembling a spacecraft or a large pipe.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a large, black, cylindrical object that resembles a spacecraft or a large pipe, positioned in a futuristic urban setting. It is the central focus of the scene, and its appearance aligns with the description. Other elements like skyscrapers and a train are present but do not contradict the core description of the cylindrical object.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a large, black, cylindrical object that resembles a spacecraft or a large pipe. situated in the futuristic city setting. The matches the central focus of the scene, and its shape aligns with the description provided The elements in thecrapers and a staircase track present but do not contradict the main description of the object object.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3a0497b7f68667a453ad61f21a8721a2330d71ea
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62bf20535495855ce33e050e7677d7f87797ebb201fb76d4c417abef9172ced5
+size 214370
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..890621ca76b9ee46a979b7380a56bfc0e702c03e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men in a sports setting, likely a stadium or arena, as they are wearing blue jackets with a logo on the left chest. They are both holding microphones to their mouths, suggesting they are commentators or announcers for a sports event. The man on the left is wearing sunglasses and has a beard, while the man on the right is wearing a baseball cap and headphones. They are both looking towards the camera, indicating they are addressing an audience. The background is a yellow wall with a metal railing, which could be part of the stadium seating or a barrier. The lighting is bright, suggesting it is daytime or the event is well-lit. The style of the video is candid and informal, capturing a moment during a sports event."
+    ],
+    "video_ids": [
+        "6t0_3iI0d4Q_9_0to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men wearing blue jackets with logos, one with sunglasses and a beard, the other with a baseball cap and headphones.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men wearing blue jackets with logos, as described. One man has sunglasses and a beard, and the other is wearing a baseball cap and headphones. These core elements are accurately represented in the video, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men wearing blue jackets with logos, one described. One man is sunglasses and a beard, and the other is wearing a baseball cap and headphones. The details elements match present represented in the video, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9cb7188b408e7cf6cd7f49c405779ba42f3b820e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae7dcb81be896a485995935ddcf610cc55932a66035032b7ad2ba4a752ddba01
+size 189321
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..6587e460b91f663d31d926454b995f96c1daeac7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a silver SOTA wheel in three different stages of rotation. The wheel is designed with a black center and silver spokes, giving it a sleek and modern appearance. The SOTA logo is prominently displayed on the center of the wheel. The wheel is shown in motion, with each frame capturing a different angle of the wheel's rotation. The style of the video is a time-lapse, showcasing the wheel's design and movement in a dynamic and engaging way."
+    ],
+    "video_ids": [
+        "PDZL0lXUL0o_4_64to207"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Silver SOTA wheel? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a silver SOTA wheel with the brand logo clearly visible in the center. The wheel's design, including its spokes and bolts, is consistent with the description. No conflicting elements are present that contradict the core description of a silver SOTA wheel.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a silver wheelOTA wheel, a logo name clearly visible in the center. The wheel's design, including its silver and the, is consistent with the description of The additional elements are present, would the core description of the silver SOTA wheel.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6ad4d4732c27aefc5e449897b968cec5ed1e4a92
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:53be63fad4b7d783c872fbb3028b06b169c822838b63ef996ed89ee0eeac46a3
+size 183006
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..e639965f190111a000940c6f09e65cadf200b2dd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures an underwater scene featuring a vibrant coral reef. The primary focus is on a large octopus, which is perched atop a cluster of yellowish-brown corals. The octopus's skin displays a mottled pattern of orange and brown hues, blending seamlessly with its surroundings. Its tentacles are extended outward, some resting on the coral while others hang loosely, showcasing the creature's flexibility and adaptability to its environment.  As the video progresses, the octopus begins to move slowly across the coral structure. Its movements are deliberate and graceful, with each tentacle carefully maneuvering over the uneven surface. The camera follows the octopus's journey, maintaining a steady focus on the animal and its interaction with the coral. The background remains consistent, with other corals and marine life visible but slightly out of focus, emphasizing the octopus's central role in the scene.  Throughout the video, the lighting is dim, typical of an underwater setting, casting a soft glow on the oct"
+    ],
+    "video_ids": [
+        "182dc25ed8703ebad650023a384f0bf159460a61068c8bfafa4668e90d669868"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Large octopus, yellowish-brown corals, other corals and marine life.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a large octopus with a reddish-orange hue, resting and moving over yellowish-brown coral formations. The background includes other coral structures and faint marine life, aligning with the description. The scene is consistent with an underwater coral reef environment, and no elements contradict the specified objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a large octopus with yellow yellowish-orange hue, which on moving on yellowish-brown cor structures. The background includes other cor and and marine hints life, whiching with the description of The vibrant is vibrant with the underwater environment reef environment, fulfilling the elements contradict the given conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..043a8c412322e0b463df058f0de8b4db2b3fa0e8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd0bca7532b58562c135518285f52e3275e9fabcb29fa2926ac7514af41f9967
+size 199148
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..80c6b296ee26808bf3cc4d95b70e0d07a35687df
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a panel discussion with a moderator and two guests. The moderator is a man with gray hair, wearing a white shirt and a microphone. He is seated in the center of the frame, facing the camera. To his left is a woman with long blonde hair, wearing a black dress. She is seated and facing the moderator. To his right is another woman with short blonde hair, wearing a floral print dress. She is seated and facing the moderator. The audience is visible in the background, seated in rows and facing the panel. The setting appears to be a television studio or a conference room. The style of the video is a standard panel discussion format, with the moderator asking questions and the guests responding."
+    ],
+    "video_ids": [
+        "kYtPezbPm4k_17_0to129"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with gray hair, a woman with long blonde hair, a woman with short blonde hair, and an audience.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with gray hair wearing a white shirt and a microphone, a woman with long blonde hair (partially visible on the left), and a woman with short blonde hair (in the middle background). The background also includes an audience, fulfilling all the described elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man with gray hair speaking a white shirt, holding woman, a woman with long blonde hair,partially visible on the left), a a woman with short blonde hair (part the foreground of). There setting includes includes an audience seated which the the conditions elements.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9d41bd8f3fc732bff80f29b0a527ece8c466fac2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5424fed0a1e84fa16a328d7722bfc4bd4b0c4940950355545b4314b54d0ebf16
+size 119312
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..70989d4467500a1fa165bbf8379bb24952122cf8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen in a motorcycle gear store, examining a red and white helmet. He is wearing a gray t-shirt with the number 38 on it. The store is filled with various motorcycle helmets and jackets, all displayed on shelves. The man seems to be carefully inspecting the helmet, possibly considering it for purchase. The overall atmosphere of the video suggests a casual and relaxed shopping experience."
+    ],
+    "video_ids": [
+        "CZ6Rww-lL34_3_31to179"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a red and white helmet, a gray t-shirt with the number 38.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man standing behind a table with a red and white helmet in front of him. He is wearing a gray t-shirt with the number 38 prominently displayed on it. The core elements described are all present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man holding in a shelf, a red and white helmet. his of him. The is wearing a gray t-shirt with the number 38 on displayed. it. The setting elements of in present present and accurately represented in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0a12ce64a5ecd60a9e8835056d709a48793758f7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1187-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee4b8ffac26290d70e2743166ecdadffa610522b6228a8e32d9cbbb158d77ba9
+size 228258
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..95e795b9f0640236a8de9bf11aa51e3fa5c96209
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a laboratory setting with a focus on a computer setup on a long wooden table. The room is well-lit with fluorescent ceiling lights, and the walls are painted a light beige color. On the left side of the frame, there are shelves filled with various equipment and supplies. A large monitor displays a graph or data visualization, which appears to be part of an experiment or simulation. The graph shows two lines labeled \"Position 1\" and \"Position 2,\" indicating some form of measurement or tracking.  In front of the monitor, there is a keyboard and a mouse placed on the table. The table also has a small device, possibly a sensor or a piece of experimental equipment, connected to the computer via cables. The camera remains stationary throughout the video, capturing the scene from a fixed angle. There are no visible characters or animals in the video, and the environment suggests a controlled, scientific setting where data collection or analysis is taking place. The overall atmosphere is one of focused activity"
+    ],
+    "video_ids": [
+        "d3fcdffb997dd68f0ff31f50d362bd2506322ac5ef34099ed8584f007840565f"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A computer setup on a long wooden table, shelves with equipment and supplies, a large monitor displaying a graph, a keyboard, a mouse, and a small device connected to the computer via cables.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a computer setup on a long wooden table, including a large monitor displaying a graph, a keyboard, and a mouse. There are also shelves with equipment and supplies visible in the background, and a small device connected to the computer via cables, as described. All core elements are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a computer setup on a long wooden table, with a large monitor displaying a graph, a keyboard, a a mouse. There are shelves shelves with equipment and supplies visible in the background. and a small device connected to the computer via cables. which described in The elements elements of present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e5e39f2e1cac8800e8a13a450738965ffc1e4176
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed290aed9830244c9c3e24f58dbdf0807c3789c3861d2082ea0136d19f85e506
+size 149503
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..dc7742569958b1a666f6386561665e4c27b840c0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young woman in a pink shirt sitting in a room with blue chairs. She is looking down and appears to be deep in thought. The room has a simple, clean design with a white wall in the background. The woman's expression is serious, suggesting that she is contemplating something important. The overall style of the video is realistic and naturalistic, with a focus on the woman's emotions and the quiet, introspective moment captured in the scene."
+    ],
+    "video_ids": [
+        "IKbcPD6G5g4_142_0to117"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman in a pink shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman wearing a pink shirt, which matches the core description. The background and other elements do not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young woman wearing a pink shirt, which matches the description description. The setting and setting elements, not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f45c04472a4dce53671590bea9436a1fba40a0c6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c59f944bb7af259849232a5c5512a1111f5eb48061a326bee142bb80a233e8d1
+size 143872
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..6dbbeb823e1f872333937e65eca366023f3f0179
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen sitting at a desk in a dimly lit room. He is wearing glasses and a white shirt, and he has his hands clasped together on the desk. In front of him, there is a lamp that is turned on, casting a soft glow on the desk. On the desk, there are papers and a glass of water. The man appears to be deep in thought, possibly working on a project or problem. The overall atmosphere of the video is quiet and focused."
+    ],
+    "video_ids": [
+        "3x1zcnKtDHg_1_0to156"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a desk, a lamp, papers, a glass of water.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man sitting at a desk with a lamp, papers, and a glass of water. These objects are present and consistent with the description, even though there are additional elements (e.g., two other people) that do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a man sitting at a desk with a lamp above papers, and a glass of water. The elements are present and match with the description. fulfilling though the is additional elements likelike.g., a glasses glasses in that are not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5529d2fabe01b99bbab24f48b6e5e40d744e4fe3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:476577207c4556c04c93a25bbf27cc56b39614ddf8253d6a7c7e50750a609fff
+size 53764
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..66f548d9d3ae2f1a5a7dd0df1b5e867c8196dbab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen interacting with a large alligator in a controlled environment. The man, wearing a white baseball cap and a gray t-shirt, is kneeling down next to the alligator, which is lying on its back on the ground. He appears to be examining or treating the alligator, possibly as part of a wildlife conservation or rehabilitation effort.  In the background, there are other people observing the scene, suggesting that this might be a public demonstration or educational event. The setting appears to be outdoors, with trees and a fence visible in the background. The overall style of the video is informative and educational, aiming to raise awareness about alligator conservation and the work being done to protect these animals."
+    ],
+    "video_ids": [
+        "FHqDjGV6YL8_3_0to181"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a large alligator, and other people observing the scene.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting on a large alligator while other people are present around them, observing or assisting. The core elements \u2014 a man, a large alligator, and other people \u2014 are clearly visible and consistent with the description. Additional elements like clothing or background details do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man interacting next the bench alligator, other people are observing in them, observing the standing. The scene elements of a man, a large alligator, and other people \u2014 are all visible and match with the description. The elements like the and the details do not contradict the main description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..74d1ed345842bf3c98dc008444ac8b3d0643f6cc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d2598ba9ffbed2492bbc7a097628c3420b8ae81bb1d94e2c9aa7cd9c2e0d5194
+size 163908
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..93d6b33fada191ffdb2e6b7ba2b39cabe333d2ee
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman standing in a kitchen, wearing a black apron with a logo on it. She has short blonde hair and is wearing a white shirt. The kitchen has a yellow wall and a wooden door. There is a window behind her, and a plant can be seen outside. The woman appears to be speaking or reacting to something, as she has her mouth open and her eyes are wide. The style of the video is casual and informal, with a focus on the woman and her immediate surroundings. The lighting is bright, and the colors are vibrant. The overall impression is that of a homey and comfortable kitchen environment."
+    ],
+    "video_ids": [
+        "rF5WiiNEBCw_14_17to193"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a black apron with a logo, short blonde hair, a white shirt, a window, and a plant.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a black apron with a logo, a white shirt, and has short blonde hair. Behind her, there is a window with a plant visible outside. All core elements from the description are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a black apron with a logo, a white shirt, and short short blonde hair. The her, there is a window with a plant visible through. The the elements of the description are present in accurately depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3996c7b8b1f62b172b91470c4600614311aa4a4c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:571d3b1c645279844cb56bfe0e6d506237b299b13764451c66ff1b2498ce9a3d
+size 138918
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..f7de3f7782350f6af50dd75a42ffb3026b18201b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the process of cooking a dish in a black skillet on a stove. The dish appears to be a stir-fry, with various ingredients being added and cooked. The ingredients include chopped vegetables and possibly some protein, such as chicken or tofu. The cooking process involves the use of oil, which is poured into the skillet and heated before the ingredients are added. The skillet is placed on a burner, which is turned on to heat the oil. The cooking process is shown in a close-up shot, allowing the viewer to see the details of the ingredients and the cooking process. The style of the video is realistic and informative, providing a clear and detailed view of the cooking process. The focus is on the cooking process itself, rather than the chef or the kitchen environment. The video is likely intended for viewers who are interested in learning how to cook a particular dish or who are looking for cooking inspiration."
+    ],
+    "video_ids": [
+        "bwCt_4EBYVk_61_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black skillet, various cooking ingredients (chopped vegetables and protein), oil, stove burner? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black skillet on a stove burner, containing chopped ingredients (likely vegetables and protein), and oil is being poured into it. The steam rising indicates cooking is underway, matching the described elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black skillet with a stove burner with with various vegetables thatveget vegetables and possibly). and oil is being added into the. This presence and and the is taking, which the description elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e5692a1e3273879f14a59c60729fdec81a3a8ad0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:316f3c9fce17bd6821f480eaddba63a038529a0dec169be088781e3186dcb9ce
+size 206221
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..7bf75a47e9b2eb7b755eaa44134f446fa69a9dfe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a black luxury car parked on a dirt road. The car is positioned at an angle to the camera, allowing a clear view of its front and side. The car's design features a large grille, sleek headlights, and alloy wheels. The background consists of a grassy area with trees, suggesting a rural or semi-rural setting. The lighting in the video is natural, with the car and its surroundings bathed in soft, diffused light. The style of the video is realistic, with no visible filters or artistic effects applied. The focus is on the car, with the surrounding environment serving as a complementary backdrop."
+    ],
+    "video_ids": [
+        "6kGnubW_xjI_0_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black luxury car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black luxury car, specifically a Mercedes-Benz, which matches the description. The car's design, color, and setting are consistent with the 'Object(s)' condition, and there are no conflicting elements that contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a black luxury car, which a Bentley-Benz, which align the description of The car is sleek, including, and features ( consistent with the 'black(s)' condition. and there are no additional elements in would the description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..12dfa27cfecf8d481df6fbfdd8c3f9a7d2da4601
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4a264dde64d5ae4fd6112f5c43a72b6af7c47f3d61f14d82e6d03789e9ffda0e
+size 253162
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..bc4f40b09fa2f31524411bfd88b992615cfbe914
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up of a pasta salad, showcasing the ingredients and their textures. The pasta is mixed with various vegetables, including red bell peppers, cucumbers, and lettuce. The vegetables are cut into small pieces, and the pasta appears to be cooked al dente. The salad is garnished with hard-boiled eggs, which are sliced and arranged on top. The eggs have a creamy yellow yolk and a white egg white. The salad is presented in a shallow bowl, and the ingredients are arranged in a visually appealing manner. The style of the video is a simple, straightforward food presentation, with a focus on the textures and colors of the ingredients. The lighting is bright and even, highlighting the freshness of the vegetables and the creamy texture of the eggs. The video does not contain any text or narration, and the focus is solely on the food."
+    ],
+    "video_ids": [
+        "lv-Rsdnh5Bc_42_0to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Pasta, red bell peppers, cucumbers, lettuce, hard-boiled eggs? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows pasta as the base, along with visible red bell pepper strips, cucumber slices, lettuce pieces, and halves of hard-boiled eggs. All the specified objects are present and identifiable in the frame.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a, the base of with with red pieces bell peppers slices, cucumber slices, and,, and hard of hard-boiled eggs. The the mentioned objects are present and identifiable in the image,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..728e32e4b2fad2148ac9f6ed3aaafab2d43cd71d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1188-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76857841e7c528c40c415ae20b1b0de4ef5567e54e79ebc9335ad30cbfb30f88
+size 85784
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..8f0644fea880b2ede1d5f75f91f2b53646235e79
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person holding a bowl of enchiladas, which are a type of Mexican dish. The enchiladas are covered in a sauce and topped with a green garnish. The person is holding the bowl over a table, which has other dishes and utensils on it. The video also includes text that provides the price of the enchiladas in pesos and US dollars, as well as the price of a lemonade. The style of the video is casual and seems to be taken in a restaurant or a home setting. The focus is on the food and the price, rather than the surroundings or the person holding the bowl."
+    ],
+    "video_ids": [
+        "AOF7hqoZD90_28_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a bowl of enchiladas, a table, other dishes and utensils? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person holding a bowl of enchiladas, placed on a table with other dishes and utensils visible in the background. The core elements described are all present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a person holding a bowl of foodiladas, which on a table. other dishes and utensils visible in the background. The presence elements of in present present and match with the description content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4cee9a299d4f032bfde8e5bfa1daf18aef64eca1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e36d2c54d0e8b326726c92eb42b9ab1dd37bfd53e33ba64266f0fb6baaf9956d
+size 127120
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..d2a457c4ec1d823518da0313a3d4c117538b0fa4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are behind a bar, preparing drinks. The man is holding a glass, while the woman is pouring a liquid from a bottle into a glass. The bar is well-stocked with various bottles and glasses. The man and woman are both wearing casual clothing, and the man is wearing glasses. The bar has a sign that reads \"CMC Common Man Cocktails\". The video captures the process of making a cocktail, with the man and woman working together to create a drink. The atmosphere is relaxed and friendly, with the focus on the art of cocktail making."
+    ],
+    "video_ids": [
+        "0KMM-XG4OvA_2_0to130"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman standing behind a bar, engaging in activities related to cocktail preparation. Both individuals are visible and identifiable as a man and a woman, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man and a woman engaging at a bar counter engaging in the that to serving preparation. The individuals are present and appear as a man and a woman based fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4c46febf8b0848241ca23758d22ea1b7a6ec0b30
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a86870d95a6655bdbee5e2cf6990d1b651f00951b5e310f3a41b5f0284314c9
+size 153474
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..1b83b6babba72c040d405972cc15179769bc1d65
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a car's dashboard and center console, focusing on the navigation system and controls. The navigation screen displays a map with a red arrow indicating the current location and a route marked in white. The controls include a large knob in the center, surrounded by buttons with labels such as \"route,\" \"map,\" \"enter,\" \"info,\" \"settings,\" and \"climate.\" The style of the video is a straightforward, static shot with no movement or action, focusing on the details of the car's interior and the navigation system. The lighting is bright, and the image is clear, allowing for a detailed view of the car's features."
+    ],
+    "video_ids": [
+        "-XbLV144cKc_26_0to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Navigation system screen, large central knob, buttons labeled 'route,' 'map,' 'enter,' 'info,' 'settings,' and 'climate'.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a navigation system screen displaying a map, a large central knob labeled 'ENTER', and buttons labeled 'ROUTE', 'MAP', 'INFO', 'SETTING', and 'CLIMATE'. The interface elements match the description, with no contradictory elements present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a navigation system screen displaying a map, a large central knob, withH', and several labeled 'route', 'MAP', 'ENTER', 'SETTINGS', and 'CLIMATE'. The presence is match the description provided and no additional elements present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6de4e814bb3446f78593e8b80898ca739541ded1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e9cc9ceaecc8127775a4ea9a02eea0728e04929939539bfd6e4ae7b3e9b06108
+size 153222
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..a5cba9ff0fe4fdaac4267d6527a3177609b0402e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a blue Volkswagen car parked on a street. The car is the main focus of the video, and it is captured from the rear view. The car is stationary, and there are no people or other vehicles visible in the video. The car's license plate is clearly visible, and it reads \"BL-44H\". The car is parked in front of a building, and there is a street lamp visible in the background. The video is a simple, straightforward shot of the car, with no additional action or movement. The style of the video is straightforward and unembellished, focusing solely on the car and its immediate surroundings."
+    ],
+    "video_ids": [
+        "JFGXvNYH5rM_0_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Blue Volkswagen car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a blue Volkswagen car, specifically a rear view of a modified sedan with custom wheels and a lowered suspension. The car's color, make, and model are consistent with the description, and no elements contradict this core identification.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a blue car car, which a model view of a blue Volkswagen. aizations and a license suspension. The car's design, model, and model are consistent with the description of and there additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d07bc137d627e5257c8763908099149fea9435f4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:904ae5bf5e06d3921ead68b223236d3e8554eaab576a7b9025b492ecf3bfac75
+size 119610
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..8bf9e60c04fcb91f02fbe76d6a8168f2fa54a07f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up shot of a car's wheel, showcasing the intricate details of the wheel's design. The wheel is black with a silver center and yellow accents, and it features the text \"GT4\" prominently. The car's brake caliper is visible, painted in a striking yellow color that contrasts with the black of the wheel. The style of the video is a high-resolution, detailed shot that focuses on the wheel, highlighting its design and the craftsmanship that went into its creation. The video is likely intended for automotive enthusiasts or for promotional purposes, showcasing the car's features and design elements."
+    ],
+    "video_ids": [
+        "AbZG3iUl5rk_14_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Car's wheel, brake caliper? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video focuses on a close-up of a car's wheel and brake caliper, clearly showing the GT4 logo on the wheel center and the yellow Porsche-branded brake caliper. These elements match the description of 'Car's wheel, brake caliper' without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows on a close-up of a car's wheel, brake caliper, which showing the details emblem logo on the wheel and and the yellow brake brakeanded brake caliper. The elements directly the description of theCar's wheel, brake caliper' without any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3e683bb4eb18605fc35289cc21597f9e8d5bec68
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6c300921d525446872e1bd377217a712eb31e36d10e62a855e60823e342fe5b
+size 276664
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..47ac87b7cea00b480a6c7833c29ce9d2ee449b9c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a white Lexus SUV in a forest setting. The car is captured from a front three-quarter angle, highlighting its sleek design and shiny chrome grille. The vehicle's headlights are on, casting a bright light on the surrounding trees and foliage. The car is parked on a dirt road, and the background reveals a serene forest with tall trees and a clear blue sky. The overall style of the video is a combination of a car advertisement and a nature documentary, emphasizing the car's design and the beauty of the natural environment."
+    ],
+    "video_ids": [
+        "wjWpEJ36Pjg_33_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white Lexus SUV? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a white Lexus SUV, identifiable by its distinctive grille and emblem. The vehicle is the central focus, and no elements contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a white vehicle SUV, which by the distinctive grille design emblem. The vehicle is the central focus, and the conflicting contradict the description.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cadcf3a949c500d28b6dbc58fca2c0f04b8eed1f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:10fc41a2498e606970df1fe11536b7c47cf6c256bbe1eb1347f542ca067c5941
+size 150756
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..7f1440255636d4faf4382a41881270b8497d2517
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a promotional advertisement for the StarCraft MX 25 DLZ boat. It features a group of people enjoying a ride on the boat across a body of water. The boat is sleek and modern, with a black and white color scheme. The passengers are seated comfortably in the boat's spacious seating area, which is equipped with cushioned seats and a canopy for shade. The boat is moving at a moderate speed, creating a wake behind it. The water is calm, reflecting the boat's hull and the clear sky above. The video is shot from a high angle, providing a comprehensive view of the boat and its passengers. The overall style of the video is dynamic and engaging, showcasing the boat's features and the passengers' enjoyment."
+    ],
+    "video_ids": [
+        "8znyup3RFh8_0_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: StarCraft MX 25 DLZ boat, group of people, cushioned seats, canopy? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the Starcraft MX 25 DLZ boat with a group of people onboard, cushioned seats, and a canopy. All core elements described in the condition are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a StarCraft MX 25 DLZ boat, a canopy of people on. cushioned seats, and a canopy. The the elements of in the condition are present in accurately depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6e7a329d84e75b40ca3bc456dbb6a9fc5ba498c7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:51ae4ee1d4150644a3ca1b2f753a8d95d1db8b2da8e69c541fbf433439b0d8e1
+size 311203
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..6550539b48f915014e101b8f096b3189f648d3c4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the breathtaking view of the Iguazu Falls, a natural wonder straddling the border of Argentina and Brazil. The first frame shows a wide shot of the falls, with the cascading water creating a misty veil that adds to the dramatic effect. The second frame zooms in on the water, revealing the powerful force of the falls as they crash down into the river below. The third frame provides a close-up view of the river, showcasing the rocky terrain and the vibrant green foliage that surrounds the falls. The video is shot from a high angle, providing a panoramic view of the falls and the surrounding landscape. The style of the video is realistic, capturing the natural beauty of the falls in all its glory."
+    ],
+    "video_ids": [
+        "WmVLcj-XKnM_1_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Water, mist, river, rocky terrain, vibrant green foliage? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts all the specified elements: water cascading down multiple waterfalls, mist rising from the falls, a river flowing below, rocky terrain along the riverbanks, and vibrant green foliage covering the cliffs and surrounding areas. The scene is consistent with the description without any conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful the the elements elements: Water cascading over the levelsfalls, mist rising from the falls, a river flowing through, rocky terrain forming the edgesbanks, and vibrant green foliage surrounding parts surrounding and surrounding areas. The animation is dynamic with the description, any contradictions elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..20a2278678a61971f05445b494b77e6eff5e7bc6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1189-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5eae03332719dd047710e9e03403ee967156221a9aa4ad5128a3be2ab62986c
+size 185664
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..2360e41c80f53ba9a277aa7e1174012d81b76db7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a woman and two children posing for a photo. The woman is standing in the middle, smiling and looking directly at the camera. The two children are standing on either side of her, also smiling and looking at the camera. They are all wearing matching blue jackets with white stripes on the sleeves. The background of the photo is a room with a blue wall and a white door. There is also a red and white tent in the background. The style of the video is a simple, candid photograph capturing a moment of joy and togetherness."
+    ],
+    "video_ids": [
+        "X19_--MOk8w_11_0to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, two children, blue jackets with white stripes on the sleeves, a red and white tent? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman and two children standing together, all wearing blue jackets with white stripes on the sleeves. Behind them is a red and white tent with the text 'A Taste for Life.' The core elements described are clearly present, and no conflicting elements are evident.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman and two children wearing together. all wearing blue jackets with white stripes on the sleeves. The them, a red and white tent, a top 'S & of Life' The scene elements of in present present, and there additional elements are introduced.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ff062d27d4f8a956c3d7928708c161d99dd45e7f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8160d0831b5d48308a10b2955af0805322882890748a8d8f55e827d86fce8a82
+size 168616
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..488c97e0184fb24bad8ba2b4c1d5a6699a46b55b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman in a floral dress, standing against a white background. She is waving her hands in the air, possibly in a gesture of greeting or excitement. The woman appears to be the main subject of the video, with no other significant objects or people present. The style of the video is simple and straightforward, focusing on the woman and her actions. The white background provides a clean and uncluttered look, allowing the viewer to focus on the woman and her actions. The video does not contain any text or additional elements, keeping the focus solely on the woman and her actions."
+    ],
+    "video_ids": [
+        "0MNekK9A2w0_0_0to117"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Woman in a floral dress? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a woman wearing a dark blue dress with a floral pattern, which matches the description 'Woman in a floral dress'. Her attire is clearly visible and consistent throughout the frames, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a dress floral floral with a floral pattern. which matches the description ofWoman in a floral dress'. The actions is the visible, the with the frames, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n addCriterion",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f875d90bd814ccba80dbaee87c7604416e41e85b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:955f0c161357f76b57153b507351158e9d125ed5c19763027e22b63acb109c1e
+size 262237
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..ad42797ba17d54decda2e39c0ca056cb7850778b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in a lush green forest. He is wearing a green shirt and a backpack, suggesting he might be on a hike or an outdoor adventure. The man is smiling and pointing at something off-camera, indicating that he is sharing a moment of joy or discovery with the viewer. The forest around him is dense with trees and foliage, creating a sense of seclusion and tranquility. The overall style of the video is naturalistic, capturing the beauty of the outdoors and the man's connection with nature."
+    ],
+    "video_ids": [
+        "Lit6M1YaczI_16_68to248"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man wearing a green shirt and a backpack, smiling and pointing off-camera.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a green shirt and a backpack, smiling, and at one point pointing off-camera. The description matches the core elements of the video, even though there are some minor variations in posture and expression.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a green shirt and a backpack, smiling, and pointing one point, off-camera. The description align the core elements of the video, and though the are no additional variations in the and the.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2850145c1662352c112480b6d4d821720e10b379
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aa46edee24cde23427265b082fae235886bca8a9dc345fdbf6a8de41176d299f
+size 204537
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..73e2c7ae1cb0e15784b4307862d08afc3504d795
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young woman with long, curly hair is seen standing against a backdrop of a wooden wall with a geometric pattern. She is wearing a white t-shirt with a graphic design on it and blue jeans. Her hands are placed on her hips, and she is looking directly at the camera with a slight smile. The overall style of the video is casual and relaxed, with a focus on the woman's attire and the unique wall behind her."
+    ],
+    "video_ids": [
+        "kJE2GA-kmoU_13_0to195"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman with long, curly hair, wearing a white t-shirt with a graphic design, and blue jeans.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman with long, curly hair, wearing a white t-shirt with a graphic design (depicting faces and text) and blue jeans. These elements match the description provided in the 'Object(s)' condition. The background and other accessories (like glasses or a hat) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman with long, curly hair, wearing a white t-shirt with a graphic design,whichicting a and a), and blue jeans. The elements match the description provided, the questionObject(s)' condition. The background is additional elements areif the) additional watch) are not contradict the description description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..83f95fe1dbb23811ec01ff1b6264e9e67b4b3951
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23927c9da6bde86b9c3aa0822b12eab6838ca1da9a867c1112dbd26e39b128ab
+size 166190
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..46fcfa59a453f13b0088779b0e7652cb529cfdf6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a joyful moment of celebration, featuring a group of people, including a young boy, gathered around a large trophy. The trophy, gleaming under the stadium lights, is held aloft by a man in a suit, who is flanked by two other men, one of whom is also wearing a suit. The group is standing on a stage, with a crowd of spectators visible in the background, their faces a blur of anticipation and excitement. The video is a snapshot of a triumphant moment, a testament to the hard work and dedication of the team. The image is filled with the vibrant energy of the stadium, the anticipation of the crowd, and the pride of the team."
+    ],
+    "video_ids": [
+        "UVt7Jo1VHV4_2_72to226"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large trophy, a young boy, three men (one in a suit holding the trophy, two others flanking him), and a crowd of spectators.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a large trophy being held by one man in a suit, flanked by two other men and a young boy. The background features a crowd of spectators, matching the description. All core elements are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a large trophy being held by a man, a suit, withanked by two other men. a young boy. The background suggests a crowd of spectators, which the description provided The elements elements are present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e0751672cc346abacf8b764301c4e02c1264b923
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:80f8f94dc0a94a22d780c7c4f69370126a147a7bd1ccd50e8a7a75f6c5401299
+size 224591
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..3d36fa5005bfe8a7f96a1c5359120b772405f9ad
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red car driving down a winding road through a lush green forest. The car is moving along the road, which is surrounded by trees and grass. The road curves gently, and the car is following the curve. The car is the only vehicle in the scene, and it is moving at a moderate speed. The forest is dense and green, with trees and bushes lining the road. The sky is overcast, and the lighting is soft and diffused. The overall style of the video is realistic and naturalistic, with a focus on the car and the surrounding environment. The video captures the beauty of the forest and the peacefulness of the road."
+    ],
+    "video_ids": [
+        "HGL1sEhpOxc_10_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red car driving along a road, which fulfills the 'Object(s): A red car' condition. Additional elements like the landscape and stacked logs do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a red car driving on a winding surrounded which matchesfills the 'Object(s)' A red car' condition. The elements such the road and the objects are not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..944858a97b9be084b50106775e50cc174c4ced9d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:422d7faadc0fac65eef3a51ad68b79742de1cd44d4224e94b185fa2a772ade0d
+size 265997
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..10bf42dc76972a43f5d22d649045464b366304c2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a room filled with yellow and gray electrical boxes. The boxes are arranged in rows, and the room appears to be well-lit. The style of the video is straightforward and documentary, focusing on the electrical boxes and their arrangement. The video does not contain any people or animals, and the focus is solely on the electrical boxes and their surroundings. The room itself is nondescript, with no distinctive features or decorations. The electrical boxes are the main subject of the video, and their arrangement and color are the primary elements of interest. The video does not contain any action or movement, and the focus is on the stillness and orderliness of the room and its contents. The video is likely intended to showcase the electrical boxes and their arrangement, possibly for educational or informational purposes."
+    ],
+    "video_ids": [
+        "GRiIxe3GfCs_87_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Yellow and gray electrical boxes? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features multiple yellow and gray electrical boxes arranged in rows, which matches the description. The boxes are clearly visible and occupy the majority of the visual space, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a yellow and gray electrical boxes arranged in a and which align the description of The boxes are the visible and dominate the majority of the frame space, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..394a9c0731d75c4b6f4f260116d8ebc15f40fc94
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:562bf84525977b8e76028f97e57a4fb5cdf79dd05ef6489080712f27dbb35fd5
+size 105306
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..33fde55aa5b4e17045dc534e91f41291d65c8890
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car from the perspective of the driver's seat. The car is a modern, dark-colored vehicle with a sleek design. The steering wheel is prominently displayed in the center of the frame, featuring a logo on the top. The dashboard is visible, with various controls and indicators. The car's door is open, revealing a window and a side mirror. The interior is well-lit, suggesting that the car is parked in a brightly lit area. The style of the video is straightforward and unembellished, focusing on the car's interior without any additional context or action."
+    ],
+    "video_ids": [
+        "AF1VbNDoKk4_24_0to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, dashboard, car door, window, side mirror? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the steering wheel, dashboard, car door, window, and side mirror as requested. All these elements are visible and accurately represented within the frame, fulfilling the 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows the interior wheel, dashboard, car door, window, and side mirror, described. The these elements are visible and match depicted in the frame, fulfilling the 'Object(s)' condition.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..44e4f6a03130f5bae93e4f7c510e9a329873ea43
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-119-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c3dcd28346f3499ec125fda4a408233bc594eebf455a11fbc846e4935d18755
+size 160544
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..2ef4c1b1a142af80d5061cce8f8f048e669f3db4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a red shirt, who appears to be in a thoughtful or contemplative state. He is looking to the side, with a slight frown on his face, suggesting he might be deep in thought or considering something important. The background is blurred, but it seems to be an indoor setting, possibly a room with a window. The lighting is soft and natural, indicating that the video might have been taken during the day. The overall style of the video is candid and unposed, capturing a moment of introspection or decision-making."
+    ],
+    "video_ids": [
+        "HM19Kvz2-V0_4_0to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a red shirt, looking to the side with a slight frown? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a red shirt, and he is looking to the side with a slight frown. His facial expression and posture match the description, and there are no elements that contradict it.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a red shirt, and he is indeed to the side with a slight frown. The expression expression and the align the description provided making there are no additional in contradict the.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e0a5dad4b5f203700236c48461e68ffffe21d0ca
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac24c8dcbf0d6b5e7cd9ddcc2e712649934c2f8186228ad62b1f9a4d73a1b1df
+size 117780
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..57f918fb582cb72485e43093f9536b847faa0e67
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a first-person perspective of a person driving a BMW convertible on a winding road. The driver's hands are on the steering wheel, which has the BMW logo in the center. The dashboard is visible, showing the speedometer and other gauges. The convertible top is down, revealing the open sky and the surrounding landscape. The road is curving to the left, and the driver is navigating the turn. The video has a cinematic quality, with a focus on the car and the driving experience. The style of the video is realistic, capturing the details of the car and the environment."
+    ],
+    "video_ids": [
+        "WljziIInekA_12_0to140"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: BMW convertible, steering wheel, dashboard, open sky, landscape? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a first-person view from inside a BMW convertible, with the driver's hands on the steering wheel and the dashboard clearly visible. The car is driving on a road surrounded by greenery, indicating an open landscape. The sky is visible through the windshield, fulfilling the 'open sky' condition. All core elements from the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a BMW-person perspective from inside a car convertible, with the steering's hands on the steering wheel and the dashboard visible visible. The open is on on a road with by a hills, which an open sky. The sky is visible through the windshield, showing the conditionopen sky' condition. The elements elements of the description are present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..72f16e3feb8ac841edb60b5d8dd736d0629a1484
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24d07d796846985589d2afabec2e020d61e549518eed30fd8be1fdf5c1f4e218
+size 230470
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..1eb758c66033e844199459f789a8395dc877e47a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a step-by-step process of assembling a salad in a mason jar. The jar is filled with layers of various ingredients, including cherry tomatoes, chickpeas, spinach, and hard-boiled eggs. The jar is placed on a wooden table, and there are additional bowls of ingredients nearby. The style of the video is a simple, straightforward demonstration of meal preparation, likely aimed at those interested in meal prepping or looking for a healthy meal option. The focus is on the jar and the ingredients, with no additional people or distractions in the frame. The lighting is bright and even, highlighting the colors of the ingredients and the texture of the wooden table. The video is likely intended to be informative and visually appealing, showcasing the process of assembling a nutritious and portable meal."
+    ],
+    "video_ids": [
+        "3bD_4IFlekY_1_0to164"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Mason jar, cherry tomatoes, chickpeas, spinach, hard-boiled eggs, wooden table, bowls? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows all the specified objects: mason jars filled with layered salad (spinach, cherry tomatoes, chickpeas, and hard-boiled eggs), a wooden table as the surface, and various bowls containing the ingredients before assembly. The visual content aligns with the described elements without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a the elements objects: aason jar, with cherry ingredients ingredientschach, chick tomatoes, chickpeas, and hard-boiled eggs), a wooden table, the surface, and bowls bowls in additional same in assembly. The addition elements aligns with the description elements without any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..63aee0c0f74702c5cdecc76e41ffe186624c18f7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe7393fbdb57e3b431822309296123df0fa71700630dc25fdff8598f64b9e15a
+size 97688
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..d096a9581e051eee683cfecf6785d29b2a9f6bd5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young man standing in front of a large white church with arched windows. He is wearing a red and black plaid shirt over a black t-shirt. His hands are outstretched, and he appears to be gesturing or explaining something. The church is surrounded by lush green bushes, and the sky is clear and blue. The man's expression is one of surprise or excitement. The overall style of the video is casual and candid, capturing a moment in the man's day."
+    ],
+    "video_ids": [
+        "BYVDvDzsc9s_62_58to203"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man, a red and black plaid shirt, a black t-shirt, his hands, the church, lush green bushes, a clear blue sky.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man wearing a red and black plaid shirt over a black t-shirt, gesturing with his hands. Behind him is a church with white stone walls and arched windows, surrounded by lush green bushes with some flowers. The sky above is clear blue. All elements described in the condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a young man wearing a red and black plaid shirt over a black t-shirt. standinguring with his hands. The him, a church with a walls walls and arched windows, surrounded by lush green bushes and a leaves. The sky is is clear and, The elements in in the condition are present in match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7fbd723cbd7764f1bd9efab777e216a360651687
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:132fa3d024b13f9b9d1b10522617cedf8462b2650b17429a5479a7ac026f3735
+size 250861
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..46abbefc7b55cf189e04c91a73acb75662d255af
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the engine of a yellow car from a top-down perspective. The engine is black and silver, with various components such as the air filter, intake manifold, and exhaust system visible. The car is parked on a gravel road, and the surrounding environment includes grass and trees. The style of the video is a close-up, real-life shot with a focus on the car's engine. The video does not contain any text or additional graphics."
+    ],
+    "video_ids": [
+        "rmprg9y-L3w_4_233to391"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A yellow car with a black and silver engine, including an air filter, intake manifold, and exhaust system.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a yellow car with its hood open, revealing a black and silver engine. Visible components include an air filter (red/purple pleated filter), intake manifold, and parts of the exhaust system. These elements match the description, and no contradictory elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a yellow car with its hood open, revealing a black and silver engine. The components include an air filter,yellow andurple),ated element), an manifold ( and parts of the exhaust system. The elements match the description provided indicating there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..988bec5c505eeb2bb3e68d5ba705ba8ea5b9d0a5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c8eb9b59f4db3335b8936a900c065864d9dcc72c907c7604b39069e5437792d
+size 209580
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..cebac2d2ad0faf9f30eab54bc72dea4e40cecd59
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse of a clock, showcasing the passage of time. The style is artistic and abstract, with a focus on the circular motion of the clock hands. The clock face is white with black numbers and hands, and the background is a gradient of blue and black, giving a sense of depth and space. The clock hands move smoothly and fluidly, creating a mesmerizing effect. The video is a visual representation of the concept of time, and the abstract nature of the image allows for multiple interpretations."
+    ],
+    "video_ids": [
+        "IsG8dfSiiTo_16_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Clock face, clock hands? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features multiple clock faces with visible clock hands, which are the core elements described in the condition. The clocks are clearly depicted, and their hands are moving, fulfilling the requirement. Additional elements like swirling blue backgrounds or other clock designs do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a clock faces with clock clock hands, which align the core elements described. the '. The presence are shown visible, and the hands are moving, indicating the requirement of The elements like the blue and do the clocks faces do not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..903f1f812db8b3747d3bc2a5dee5837505129d51
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:734a1b15f99ab4195ea5bc487b7ddb1a1a33aa8a23785197d5f43ae50fbbada5
+size 186414
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..e54bf53d436bc82ba534c82295a229ee49137d54
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man driving a car with a snowy mountain landscape in the background. The man is wearing a blue t-shirt and a seatbelt. He is holding a steering wheel with both hands. The car has a black interior. The man appears to be speaking into a microphone. The video is shot from the perspective of the passenger seat, giving a first-person view of the drive. The style of the video is realistic and it captures the beauty of the mountain landscape."
+    ],
+    "video_ids": [
+        "DkVI3FLQZAI_8_78to251"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a car, a microphone? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man driving a car, with a microphone visible attached to his shirt. All three specified objects are present and accurately depicted in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man driving a car, which a microphone visible on to his shirt. The the elements objects ( present and match depicted in the scene.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..af7588335d96a14ec2cb8091ae05b1fc5250c631
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3f2c53bfe102754685a2d72b9782e7eb1836b8f5c91d3227c03ad7f443f566e8
+size 234059
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..349c7a04e2acfb01afa5f7128d8c8ce03c373f3b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a white speedboat cutting through the water at high speed. The boat is sleek and modern, with a curved windshield and a large cabin. The water around the boat is churned up into a frothy white wake, indicating the boat's speed. The boat is moving towards the right side of the frame, and the water appears to be a deep blue color. The sky is overcast, and there are no other boats or objects in the water. The style of the video is dynamic and action-oriented, with a focus on the speed and power of the boat."
+    ],
+    "video_ids": [
+        "eSP0sWFI2cI_9_0to151"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white speedboat with a curved windshield and a large cabin.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white yacht with a curved windshield and a large cabin, matching the described object. The yacht is moving at speed, creating a wake, which is consistent with a speedboat. Additional elements like the flag and the ocean background do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white speed with a curved windshield and a large cabin, which the description object. The yacht is in at high, creating a wake, which is consistent with the speedboat. The elements like the ocean and the ocean background do not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e6a9dfa6ad1709bcd09403441ca42fea5ccaca31
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1190-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e1cac782738f099c98b9632e1a84485264ba5f4c6f76645b69170a760b8087e1
+size 225127
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..8adcd4b4bb0a2a3e3770d9a126382dbeb38f2def
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of two boats sailing on a clear blue body of water. The boats, one larger and one smaller, are moving in the same direction, leaving a trail of ripples in their wake. The larger boat is closer to the camera, while the smaller boat is further away, both appearing to be moving towards the right side of the frame. The water is a vibrant blue, reflecting the clear sky above. In the distance, a small island can be seen, adding a touch of land to the otherwise expansive water body. The overall style of the video is calm and peaceful, capturing the beauty of a day out on the water."
+    ],
+    "video_ids": [
+        "AG-TYpMi6Ek_27_743to937"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two boats (one larger, one smaller), water, sky, island? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two boats \u2014 one larger sailboat and one smaller motorboat \u2014 on turquoise water, with a visible island in the background and a clear sky above. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts two boats, one larger andboat and one smaller motorboat \u2014 traveling a water. with a small island in the background. a sky sky above. The the elements of in present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9aaed7543565fdc283f69717e27de3d9db0014e5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b8367b5f1ef0acbdfb43d6cfe568e1c52a3982269198628b658d0f2084f44c3
+size 233036
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..e49c441c3752abfd925be77f07f94c868deef0dd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and glasses, wearing a pink shirt. He is standing in a room with a brick wall and a lamp in the background. The man appears to be speaking or presenting, as suggested by the presence of a microphone. The style of the video is casual and informal, with a focus on the man and his surroundings. The lighting in the room is soft and warm, creating a comfortable atmosphere. The brick wall and lamp add a touch of rustic charm to the setting. The microphone suggests that the man may be recording a podcast or conducting an interview. Overall, the video seems to be a personal or professional project, possibly related to the man's interests or expertise."
+    ],
+    "video_ids": [
+        "a3qbqXRyFMk_0_0to186"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and glasses, wearing a pink shirt, standing and speaking or presenting. A microphone is present.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and glasses, wearing a pink shirt, who is speaking or presenting. A microphone is visible clipped to his shirt. The background and other elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and glasses, wearing a pink shirt, standing appears standing or presenting. A microphone is visible in to his shirt, The background includes setting elements in not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..da175b7ef18f5b45d47985b6df0a17367da3fc0d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:25060b10f87829840fc0af77ebf0dd0f51c8206b42e2951566f3129cd98d08f5
+size 121888
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..1dfc78a436079ed97e12a52c6fc55a74ea27d708
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a brown, furry animal, possibly a badger or a similar creature, walking through a forested area. The animal is seen in three different frames, each showing it in motion, with its tongue sticking out in the last frame. The forest around the animal is lush and green, with various trees and foliage visible. The animal's fur appears to be well-groomed and healthy, suggesting that it is in good condition. The video is shot in a realistic style, with natural lighting and colors that accurately represent the scene. The animal's movements are captured in a way that gives a sense of its size and strength. The overall impression is of a wild, untamed creature in its natural habitat."
+    ],
+    "video_ids": [
+        "Dvf5bLjR_KA_19_0to128"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A brown, furry animal, possibly a badger or a similar creature? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a brown, furry animal with a robust build, small ears, and a snout that resembles a badger or a similar creature. Its fur color and body shape are consistent with the description, and there are no elements that contradict this characterization.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a brown, furry animal that a distinctive build and which ears, and a shortout, is a badger. a similar creature. The fur is and texture structure align consistent with the description of and it are no elements in contradict this description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7bb5f1f55df39f89b4157d73a069a7ef46c5f118
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:df710a2566e538ec1a3bf83e20bc87b9fa19d06d89ef9c6734464ee0e7862f28
+size 207569
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..fab01e4bbe0f3037b9a6a3e1bfc187528ca20930
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a conversation between two men in a professional setting. The man in the foreground is dressed in a suit and tie, with a green patterned tie, and appears to be speaking. The man in the background is wearing a vest and a white shirt, and is facing the man in the suit. Both men are standing and appear to be engaged in a serious discussion. The setting is an office with a beige wall and a window in the background. The lighting is bright and even, suggesting an indoor environment. The style of the video is realistic and appears to be a scene from a television show or movie."
+    ],
+    "video_ids": [
+        "1QDxzGh41_U_6_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, one in a suit with a green patterned tie, and another in a vest and white shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men engaged in conversation. One man is wearing a suit with a green patterned tie, and the other is wearing a vest over a white shirt. These details match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two men. in a. One man is wearing a suit with a green patterned tie, and the other is wearing a vest and a white shirt. The details match the description provided, the questionObject(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e5f80d2f9679a58269e618247f79bb31db661f9e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0cf3c2587f30701d42ed1bfe1ad90172b3882c6cce4f178dd7a16ca5c70b8af6
+size 185998
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..635b23aa607d3e19c924d6a461636c0e36b0a615
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features an older man with a mustache and glasses, standing in front of a vintage car. He is wearing a gray jacket and appears to be speaking or presenting. The car is a classic model, possibly a convertible, and is parked on a dirt road. The setting appears to be an outdoor area with a brick building in the background. The man's expression is serious, and he seems to be addressing the camera or an audience. The overall style of the video suggests it could be a documentary or a news segment, focusing on the man and the car."
+    ],
+    "video_ids": [
+        "3B0XGs1fdrk_23_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An older man with a mustache and glasses, wearing a gray jacket, and a vintage car (possibly a convertible).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an older man with a mustache and glasses, wearing a gray jacket, standing in front of a vintage car (which appears to be a light green convertible). The description matches the core elements of the video without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features an older man with a mustache and glasses, wearing a gray jacket, standing in front of a vintage car thatpossibly appears to be a convertible-colored convertible). The setting matches the core elements of the video, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5e32fc9f950dfa4e528da09c82b131472026a9b3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e78ef488104e14b3b3caf974a5256df651616995bc895bd899b3772a57352c04
+size 180126
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..23c3a95741881797dc9f1efc72d5e3a2b9a2281b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man piloting a small aircraft, wearing a blue shirt and sunglasses, with a headset on. The aircraft is flying over a lush green field with a runway visible in the distance. The sky is partly cloudy, suggesting a pleasant day for flying. The man appears focused and attentive, indicating that he is actively engaged in piloting the aircraft. The overall style of the video is realistic, with a focus on the man and his surroundings, providing a sense of the experience of flying a small aircraft."
+    ],
+    "video_ids": [
+        "BrTJPIGb26Y_15_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue shirt and sunglasses, wearing a headset, piloting a small aircraft.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a blue shirt and sunglasses, wearing a headset, seated in the cockpit of a small aircraft. He appears to be piloting the aircraft, as indicated by his position and the view of the runway outside. The description is largely accurate, with no significant contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue shirt and sunglasses, wearing a headset, who in the cockpit of a small aircraft. The appears to be piloting the aircraft, as indicated by his position and the presence of the cockpit and the The presence matches largely matched, with the significant contradictions or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6abbb266794b71519bd29a628180905c5b404f4d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5fcf4ad9c8b3f88dfbff2a9014242fa6eaadaebc567234fa907f7927b30e1058
+size 223660
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..2b83195b3cf559d140d6cf1eff266a8ab3c37112
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a journey on a winding road through a rural landscape. The car's side window is open, revealing a clear blue sky and a lush green field. The road curves gently, and the car's side mirror reflects the surrounding scenery. The car is moving at a moderate speed, and the driver's hand is visible on the steering wheel. The overall style of the video is a realistic depiction of a road trip, with a focus on the natural beauty of the countryside."
+    ],
+    "video_ids": [
+        "DqVsTrXNKCo_31_0to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A car, a clear blue sky, a lush green field, and the car's side mirror.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a car (visible from the side), a clear blue sky with some clouds, a lush green field alongside the road, and the car's side mirror (visible on the passenger side). All core elements of the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a car inpart in the side mirror a clear blue sky, a clouds, a lush green field, the road, and the car's side mirror reflectingreflect in the left side). The these elements of the description are present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..604ee688ea70dad784a8db43881ac708d5e15275
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6f149d61c24e2f097b4bda86f97c759044b11fdf3175e760c4043d65f2ead9ab
+size 274452
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..6368845101940ede19b6b4f101977c81c7d064bc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, three young men are seen in a room with a vibrant, colorful background. The room is filled with various objects, including a TV, a lamp, and a couch. The men are dressed casually, with one wearing a green and white plaid shirt, another in a black sweater, and the third in a black sweatshirt with colorful text. They are engaged in a conversation, with one man gesturing as he speaks. The overall style of the video is casual and relaxed, capturing a moment of camaraderie among friends."
+    ],
+    "video_ids": [
+        "RzoTHsEhgZw_8_0to174"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three young men, a TV, a lamp, and a couch.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows three young men standing in the foreground. In the background, a TV with a glowing screen is visible on the left, a lamp is on the right side of the frame, and a couch can be seen behind the men. All specified objects are present and do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows shows three young men standing in a foreground, In the background, there colorful is a colorful screen is visible, the left side and lamp is hanging the right,, the frame, and a couch is be seen partially the men. The elements objects are present, match not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b5ec93fab1fe9244a101dfea7288edfd8c6b92b0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1191-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7c8c274c6b94a67486876e34bad71b8d50c7f007e5f29496316d7ab208960a60
+size 296533
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..55a8dc5e807e3f8b9a9b57178070b9a253987035
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a drone shot of a small island with a church and a bell tower. The island is surrounded by water and is covered in trees. The church and bell tower are the main focus of the video. The drone flies over the island, providing a bird's eye view of the church and bell tower. The video is likely to be a travel or nature documentary, showcasing the beauty of the island and its unique architecture. The drone's perspective allows for a comprehensive view of the island, highlighting its isolation and tranquility. The video is likely to be informative, providing viewers with a glimpse into the island's history and culture."
+    ],
+    "video_ids": [
+        "Ussi-y9v5h4_10_44to231"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Church, bell tower? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a church with a bell tower situated on a small island surrounded by water, matching the 'Object(s)' condition. The church and bell tower are central to the visual content, and no elements contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a church with a bell tower, on an small island surrounded by water. which the descriptionObject(s)' condition. The church has bell tower are the to the scene content, and there additional contradict this description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ecc5b0e5bda7ec86dbcdecf9889729cc1a1d6f0c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:10ea2ffd722faff0c4bf82ae76d7faddf44cc0e3351fc85d59ef66013df6c05a
+size 194449
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..6306e6effd10e6810beee3990a61171258192447
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, an elderly woman and a young man are seated at a wooden table, enjoying a meal together. The woman, dressed in a vibrant red jacket, is seen using a spoon to eat from her bowl. The man, wearing a blue shirt, is also eating from his bowl. The table is set with two bowls, each containing a meal. The background features a brick wall, adding a rustic charm to the scene. The video captures a heartwarming moment of intergenerational bonding over a shared meal."
+    ],
+    "video_ids": [
+        "03ucHF-TN5Y_38_22to194"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An elderly woman and a young man.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows an elderly woman with gray hair and glasses, wearing a red jacket, and a young man with brown hair, wearing a blue sweater and scarf. Both are seated at a wooden table eating food from bowls, which matches the description of 'An elderly woman and a young man.'\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows shows two elderly woman and long hair and a, wearing a red top, and a young man with short hair, wearing a blue shirt. glasses. Both are seated at a table table, from from bowls. which align the description of anAn elderly woman and a young man.' The}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1517dc9f5f88a7326b4bb300f0597ef8cd941864
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9d6a238aa43b42c496f59121868577500b82c892ff8da4d29749b4c59b621754
+size 204464
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..bb1710cafe4fb6e6021959ca7eb2621b8a540272
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a black car's wheel and tire. The tire is a standard black rubber with a visible tread pattern. The wheel is a silver alloy with a multi-spoke design, featuring a shiny metallic finish. The car's bodywork is also black, and the wheel is positioned at the front of the vehicle. The style of the video is a straightforward, static shot with no movement or action. The focus is solely on the wheel and tire, with no additional context or background provided. The lighting in the image is bright, suggesting it was taken during the day."
+    ],
+    "video_ids": [
+        "kN__dhihmTE_105_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black car's wheel and tire? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a close-up of a black car's wheel and tire, including the alloy rim, tire tread, and brake components. The car's body is also visible, confirming it is a black car. There are no elements that contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows features a car-up of a car car's wheel and tire. which the tire rim and tire tread, and part cal. The tire appears body is not partially, confirming that is black black car. The are no additional in contradict the description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0fa6a6cb5c03a4f0c9d2f693d6b3f2973323e03d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5dde0c745eb47bc5fcccca92d9af3d761c934bdd565d64f8f5cb0aee11ac64be
+size 118890
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..2b5cb2aee55ac97021b9af05fc1b4babbf6660d3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red and black Ducati motorcycle parked in a parking lot. The motorcycle is positioned in the center of the frame, with its front facing the camera. The background features a variety of cars parked in the lot, and a building can be seen in the distance. The lighting suggests it is daytime. The style of the video is straightforward and documentary, capturing the motorcycle in its environment without any additional action or movement."
+    ],
+    "video_ids": [
+        "sU6XpnKvFmA_14_206to381"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red and black Ducati motorcycle parked in the center of the frame.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red and black Ducati motorcycle parked in the center of the frame, with clear visibility of its design and position. While there are other elements in the background (cars, buildings, trees), they do not contradict the core description and are consistent with a typical parking lot setting.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a red and black Ducati motorcycle in in the center of the frame. which the visibility of its design and color. The there are additional elements like the background,such and trees, and), they do not detr the main description of serve secondary with the typical urban lot setting.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b533ca481c6a5734c859a4c8ac8e387c498b7b25
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef4029eab3b922858c748c7717124e5543145750fe25b2cfe0843939ed14a291
+size 175078
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..3619d8646bc0ad4d0b3904801b0c8cf36426ad66
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a basketball game. The main focus is on two players, one in a white jersey and the other in a blue jersey. The player in the white jersey is in possession of the ball, dribbling it towards the basket. The player in the blue jersey is in a defensive stance, ready to intercept or block the shot. The background is filled with the blurred figures of other players and spectators, indicating that the game is taking place in a crowded arena. The style of the video is a fast-paced, action-packed sports sequence, capturing the intensity and excitement of the game."
+    ],
+    "video_ids": [
+        "gpL3rQQ7KCs_0_31to201"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two players (one in white jersey, one in blue jersey), ball, basket, blurred figures of other players, spectators.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two basketball players\u2014one in a white jersey and one in a blue jersey\u2014engaged in a defensive play. The player in white is holding the ball, and the player in blue is defending. Blurred figures of other players and spectators are visible in the background, matching the description. There is no visible basket in the frame, but since the description allows for additional elements as long as they don\u2019t conflict with the core description, this omission is acceptable.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts two basketball players, in a white jersey and one in a blue jersey\u2014engaged in a game play. The player in the is attempting the ball and preparing the player in blue is attempting. Theurred figures of other players and spectators are visible in the background, which the description. The are also significant conflict in the video, but the the players does for the elements, long as they do't conflict with the core description, the is is acceptable.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fd3b0f32bcd6bf729cbfc6a3e91bbe530a330daf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c8e73ddd6de9eed2ac679c90365bf94a0feb7d1041c6114a6806e76ca9365e52
+size 277514
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..9f9c4869d85572e39f1efbbc1c65d2d2a4764801
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated scene featuring three dolls in a donut shop. The first doll, a girl with blonde hair, is standing in front of a display case filled with various donuts. She is wearing a blue dress with pink polka dots and has a pink bow in her hair. The second doll, a girl with brown hair, is standing behind the display case. She is wearing a blue apron and is holding a tray of donuts. The third doll, a girl with blonde hair, is standing in front of the display case. She is wearing a pink dress with a white collar and has a pink bow in her hair. The donut shop has a yellow and pink color scheme and features a sign that reads \"Missy Donut\". The dolls are the main focus of the scene, and their actions suggest that they are either customers or employees of the donut shop."
+    ],
+    "video_ids": [
+        "Tag87oS1PFY_75_49to191"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three dolls, two with blonde hair and one with brown hair.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows three dolls: two with blonde hair (one with a high ponytail and one with a side ponytail) and one with brown hair standing behind the counter. All three dolls are visible and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows three dolls, two with blonde hair andone in a pink ponytail and the with a low ponytail) and one with brown hair. behind a counter. The three dolls are positioned and match the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..98274c410ba855f97a0a4f6fcbe55de2666aa628
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6e0e8d1eb641bca45613d05ef87235b27a29b856e954174073daeebb386f1569
+size 137354
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..ad4ea2b70c3c40ee52c863a19689495b7dde84b6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of the interior of a car, focusing on the steering wheel and dashboard. The steering wheel is prominently displayed in the center of the frame, with the car's logo in the center. The dashboard is sleek and modern, featuring a digital display and various controls. The car's interior is well-lit, with the dashboard and steering wheel reflecting the light. The style of the video is realistic and detailed, capturing the textures and colors of the car's interior with precision. The video is likely intended for promotional purposes, showcasing the car's interior design and features."
+    ],
+    "video_ids": [
+        "0Sq1lr0CORk_19_0to145"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, car's logo, dashboard, digital display, various controls? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the steering wheel with the car's logo (Volkswagen emblem) prominently displayed. The dashboard is visible, including a digital display showing speed and other metrics. Various controls are also visible, such as the gear shift lever, climate control knobs, and buttons on the steering wheel. All specified elements are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a steering wheel, the car's logo inaolkswagen),) in displayed in The dashboard and visible, and the digital display showing the and other information. Various controls, also visible on such as the horn shift and and air control buttons, and other on the steering wheel. The these elements are present and match depicted,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..aac521dc86be31e23c1b08088033835524877a61
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dcd7d3e6440b7a3d86ad142636f499474f8cc08d70d7c5d6689c39120728b666
+size 177378
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..3b3fa8859193f5684d2f879ede6be99eac71256c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man with a beard and tattoos on his arms, speaking in a room with a stone wall. The man is the main subject of the video, and he is the only person visible in the frames. The style of the video is casual and informal, with a focus on the man's facial expressions and body language as he speaks. The room has a simple and uncluttered design, with a white wall in the background and a couch visible in the lower frame. The lighting in the room is soft and natural, suggesting that the video was taken during the day. The man's tattoos are visible on his arms, adding a personal touch to the video. The overall impression is that of a candid and spontaneous moment captured on camera."
+    ],
+    "video_ids": [
+        "oC-EpKREd2U_5_25to156"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and tattoos on his arms.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man with a beard and visible tattoos on his arms, which matches the described condition. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and visible tattoos on his arms, which align the description '. The man and setting elements in not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bc92338040b30beb2f665a5ebf2d70b2ced1eac3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1192-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c1a6d64f1e736e6a685da01e61e625448929566ab85f8416c8a8107c54e70be
+size 212346
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..bdc867ff6960bd001bd838d9a734afd314ac9dc0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit and tie, sitting in front of a news backdrop. He appears to be a news anchor or reporter, as he is speaking into a microphone. The backdrop includes a logo with the number \"5\" and the text \"POLICE OVERTIME PAY RACKING UP AFTER BIG EVENTS\". The man's expression is serious, suggesting that he is discussing an important topic. The overall style of the video is professional and newsworthy."
+    ],
+    "video_ids": [
+        "SYpkISg-R_Y_3_75to270"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, a news backdrop with a logo, a microphone? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit and tie, positioned in front of a news studio backdrop that includes a prominent logo (the number '5'). A small microphone is visible near his collar, confirming the presence of all specified elements. Additional graphics (like social media icons and news tickers) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man dressed a suit and tie, which in front of a news backdrop backdrop that includes a logo logo.likely ' '50 There microphone microphone is visible in the mouth, indicating the presence of a the elements. The elements ande the media handles) text-relateders) are not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..420d63a0f24baccd3ccb8b765734e80e3c4068c9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60ab8b6b9584949501bca8fe139e296110e3aad01da13efeba30bea8ee5e0388
+size 153005
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..2f416a548b7339f937312f74822cfdadf4b430a2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a yellow plastic toy designed to resemble a fish. The toy is placed on a rough, textured surface, likely concrete or asphalt, which appears slightly worn and scattered with small pebbles. Inside the toy's open mouth, there are several colorful, round objects that resemble small toys or balls, each with distinct patterns and colors such as blue, green, pink, and purple. The camera remains stationary throughout the sequence, focusing on the toy and its contents. There are no visible characters or animals interacting with the scene. The lighting suggests an outdoor setting during daylight hours, with natural light illuminating the toy and its surroundings. The video does not show any significant changes or movement, maintaining a consistent focus on the toy and its vibrant, multicolored contents."
+    ],
+    "video_ids": [
+        "18a4cf6ede017277718366bbdff7e87d32745261fe029bca4b3007e7a5ed11cb"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Yellow plastic toy resembling a fish, several colorful, round objects inside the toy\u2019s mouth? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a yellow plastic toy shaped like a fish with several colorful, round objects inside its mouth-like opening. The toy's design and contents match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a yellow plastic toy that like a fish with its colorful, round objects inside its mouth. opening. The objects is design and the match the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9704b73a9d425e67fa8d569d5950913592192890
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f727ecf9c45711ebc42ed32905859302c6a1aea5d4fb2e863af6d03b57bca3dc
+size 191190
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..f2cf5a14741c05678487c46976ca092c14b31a6a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen enjoying a drink from a coconut. He is wearing a white t-shirt and a backpack. The coconut is being held by another man who is wearing a green t-shirt. The scene takes place in a market, with various fruits and vegetables visible in the background. The man in the white t-shirt is seen taking a sip from the coconut, while the man in the green t-shirt is holding it steady. The overall atmosphere of the video is casual and relaxed, capturing a moment of enjoyment in a bustling market setting."
+    ],
+    "video_ids": [
+        "OCS8KhakJhY_109_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a white t-shirt, a man in a green t-shirt, and a coconut.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a white t-shirt drinking from a coconut with a straw, and another man in a green t-shirt holding the coconut. These core elements match the description exactly, and no contradictory elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man in a white t-shirt and from a coconut, a straw, and another person in a green t-shirt holding the coconut. The elements elements match the description of, fulfilling there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..faf68ba2522b3b83e9b525039acc2ae4703a4cfc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c3aca2f0e307ffc029afcc86ba396782486bed49b62dc70992f5e3e094b30ef
+size 170052
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..a136f430098f3228d1bc426311da9b076ed942ba
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are sitting on a white couch in a cozy living room. The man is holding a remote control and appears to be watching something on a television that is not visible in the image. The woman is wearing a red sweater and is holding a small elf figurine. She is looking at the man with a playful expression on her face. The room is decorated with various Christmas decorations, including a snowman and a reindeer. The overall atmosphere of the video is warm and festive."
+    ],
+    "video_ids": [
+        "HTW2QR8GUUo_1_0to147"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a woman, a remote control, a television, a red sweater, and a small elf figurine.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a man and a woman sitting on a couch, with the man holding a remote control. The woman is wearing a red sweater, and a small elf figurine is visible on the couch in front of them. Although a television is not directly visible, the presence of a remote control implies it is nearby, which is consistent with the context. All core objects mentioned are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a man and a woman sitting together a couch, both the man holding a remote control. The woman is wearing a red sweater and and there small elf figurine is visible in the table. the of them. The a television is not directly visible, the context of the remote control suggests that is likely, fulfilling is a with the ' of The the elements are in present or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2fca414d5212bd574c297a66e90adf9f6e1ec1ff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a9424e6cc4d2e256269db7b6bb051c4a6860a60f9a3ca00ad0da64401a50ec1
+size 186799
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..61afb0d3d5feb3239e7aefdb2e384ae8ef98eb39
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a vintage car, focusing on the back seat. The car's interior is green and white, with a shiny finish on the seats and door panels. The back seat is empty, and the car appears to be parked in a driveway. The car's design suggests it is from the 1950s or 1960s. The video is a still image, and there is no movement or action. The style of the video is simple and straightforward, focusing on the car's interior without any additional context or background."
+    ],
+    "video_ids": [
+        "xmUJixqoUQM_15_0to135"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Back seat (empty)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows the back seat of a car, and it is clearly empty with no people or objects on it. The focus remains on the seat's design and upholstery, which aligns with the description of an empty back seat.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows the interior seat of a car, which it is indeed empty. no passengers or objects present it. The focus is on the seat, design and condition, which aligns with the description of an empty back seat.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fd2f6d97a76ecb68e237b41c4fcb8fb3b924923f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9303d1e2c7ac17bf6ac8a68539894a0de800ccf5d502be1c7f3c651c45c93f47
+size 74364
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..7393dea34830415dc6ef5f5696a821cae8397972
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a brown bear is seen in a kitchen setting, interacting with various objects. The bear is holding a magnifying glass over a pink chair, as if inspecting it. On the table in front of the bear, there are several fruits, including apples, oranges, and a banana. The bear appears to be curious about these items. In the background, there is a refrigerator and a window with curtains. The overall style of the video is cartoonish and whimsical, with a focus on the bear's actions and the objects in the scene."
+    ],
+    "video_ids": [
+        "HNad9oDp0uY_219_75to269"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Brown bear, magnifying glass, pink chair, fruits (apples, oranges, banana), refrigerator, window with curtains? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a brown bear holding a yellow cup (which may be mistaken for a magnifying glass due to its shape but is clearly a cup), sitting in front of a pink chair. On the table, there are fruits including apples, oranges, and a banana. A refrigerator and a window with curtains are also visible in the background. All core elements from the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features ful a brown bear holding a magn magn,which could be a for a magnifying glass due to its shape), is not not cup in sitting on a of a pink chair, There the table, there are fruits including apples, oranges, and a banana. The refrigerator and a window with curtains are also visible in the background. The these elements from the description are present in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..02d2d60322f0045f5b6c0fe097e12e8e2fd4d447
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37e161b8a2550f6f350e097d6103324a1a6dea06cf68c187d69c838f3d671d51
+size 151704
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..306325c6510cd80e849124c46ec52835a55f6774
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features an elderly woman with white hair, wearing a yellow jacket and a pink scarf. She is smiling and appears to be in a good mood. The background consists of a blurred image of red flowers, which suggests that the setting might be a garden or a park. The overall style of the video is warm and inviting, with a focus on the woman's joyful expression and the vibrant colors of the flowers. The video seems to be a snapshot of a pleasant moment in the woman's life, capturing her happiness and the beauty of nature around her."
+    ],
+    "video_ids": [
+        "L9xL1orfAqo_71_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An elderly woman with white hair, wearing a yellow jacket and a pink scarf, smiling.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an elderly woman with white hair, wearing a yellow jacket and a pink scarf, and she is smiling. These elements match the description provided in the 'Object(s)' condition. The presence of Korean text on the right side of the screen does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts an elderly woman with white hair, wearing a yellow jacket and a pink scarf, and she is smiling. The elements match the description provided. the questionObject(s)' condition. The background of red text in the jacket side of the image does not contradict the description description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..644e92d2c4b4e29a5a4b610e3447ecf5baf3804c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af0b4f8110aa70686cb54ea597fa2518e301b928942b4c28c3bc2d5df8e9459f
+size 227193
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..dadff74f77e4c83a9b279641c1e81bccb98f543b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with white hair and glasses, wearing a blue jacket, engaged in a conversation with another woman whose back is facing the camera. The setting appears to be an indoor space with orange walls and a large window in the background. The woman with white hair is speaking and appears to be listening attentively to the other woman. The overall style of the video is a candid, real-life interaction between two individuals, captured in a natural setting."
+    ],
+    "video_ids": [
+        "MlIvag8Xi2s_16_0to148"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two women - one with white hair and glasses wearing a blue jacket, the other whose back is facing the camera.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two women. The woman on the left has white hair, wears glasses, and a blue jacket, matching the description. The woman on the right has her back to the camera, also matching the description. There are no elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two women. One woman with the left has white hair and is glasses, and is blue jacket, which the description. The woman on the right has her back to the camera, which align the description. The are no additional in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d945e4f8607598b3eeaa4fcc51e23e3a23089fa4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1193-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f9bf99a57a7c7c55398f8f6122b876155571c42121f4a626fdfdc18ddddd0b12
+size 163203
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..fbac058e92290a85bc06d99e5001b6277210f67f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a silver SUV parked on a dirt road in a desert-like environment. The vehicle has a large, open-top tent attached to its roof, which is supported by a ladder. The sky is cloudy, and the sun is setting, casting a warm glow over the scene. In the background, there are hills and a few scattered bushes. The style of the video is realistic, with a focus on the vehicle and its surroundings. The camera angle is from a low perspective, looking up at the vehicle and the tent. The video captures the sense of adventure and freedom associated with outdoor activities and camping."
+    ],
+    "video_ids": [
+        "3D2NZKd4lxE_7_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Silver SUV, large open-top tent on the roof, ladder supporting the tent, cloudy sky, setting sun, hills, bushes? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a silver SUV with a large open-top tent mounted on its roof, supported by a ladder. The background features a cloudy sky with the setting sun, and the landscape includes hills and sparse bushes, matching the described conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a silver SUV with a large open-top tent on on its roof. supported by a ladder. The background features a cloudy sky, a setting sun, and there landscape includes hills and bushes bushes, which the description elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..051e0acead338625d9e575f99f23b4aa4ea864f6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48cd6708676040e1d9ba2c1f7be80bbcefdad0816c5e4ba2e95efdbdedf9dd7b
+size 210805
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..137a413ca5e5d4a30925538619e9521e040abeb6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man wearing a black face mask, standing in front of a fence on a rainy day. The man is wearing a blue t-shirt and has short blonde hair. He is looking directly at the camera with a slight smile. The background features a wet street with puddles and a traffic light. The overall style of the video is casual and candid, capturing a moment in the man's day."
+    ],
+    "video_ids": [
+        "O6RT_ZFrOsc_0_0to180"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man wearing a black face mask, a blue t-shirt, and short blonde hair.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with short blonde hair, wearing a black face mask and a blue t-shirt, which matches the core description. Additional elements like the wet pavement, traffic, and background scenery do not contradict the description and are acceptable as they provide context to the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing short blonde hair wearing wearing a black face mask, a blue t-shirt. which matches the description description provided The elements such the background ground and green light and green trees do not contradict the main and are acceptable.\"\n long do context to the setting.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..be091c5697878b7b6e369d0dec13936eb5dd1d2c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e731f80eb86ea7aa13ad55214fc6e5591c7d45fe447c6f4af102f841b6784de
+size 125269
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..d27b7ac95534f381febc516fda7fda771626c3a1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a hand is seen holding a red and black electric drill, which is being used to drill into a yellow foam material. The foam material is placed on a white surface, and there are several other cups of the same material scattered around the area. The hand holding the drill is positioned in such a way that it appears to be drilling into the foam material. The drill is being used to create holes in the foam material, which is likely being used for a craft or DIY project. The overall style of the video is simple and straightforward, focusing on the action of drilling into the foam material."
+    ],
+    "video_ids": [
+        "6lLgQsf3hp4_9_798to952"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: hand, red and black electric drill, yellow foam material, white surface, cups? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a hand holding a cup filled with yellow foam material, and a red and black electric drill is being used to interact with the foam. The scene is set on a white surface, and there are additional cups visible in the background. All elements mentioned in the condition are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a hand holding a red, with a foam material, which a red and black electric drill is being used to drill with the foam. The white is set on a white surface, which the are additional yellow with in the background, The the described in the condition are present and match with the video.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a4ec690f45eb8c8e03e8b3a57c8f699794af73c0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:949f4b1736b9367132bdbb217933f032761a8b303a8bf3c3f4fa36fe9217affa
+size 148340
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..693d4c768ebc12b719ebfc4debd8586f753b06b1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows three brown balls with specks of orange and yellow on them, placed on a white surface. The balls appear to be made of a textured material, possibly a type of food or a decorative item. The balls are arranged in a triangular formation, with the largest ball at the front and the smallest ball at the back. The balls are stationary and do not interact with each other or the surface they are placed on. The style of the video is simple and straightforward, focusing on the balls without any additional context or background elements. The lighting is even, and there are no shadows or reflections visible on the surface. The video does not contain any text or sound."
+    ],
+    "video_ids": [
+        "t-wZnisZzTc_7_0to117"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three brown balls with specks of orange and yellow.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three round, brown objects that appear to be balls or truffles. They are textured with visible specks of orange and yellow, matching the description. The background is neutral and does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three spherical objects brown objects with are to be balls. spheresink. They are covered with spe specks of orange and yellow, which the description. The objects is plain, does not distract the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a97a00328d468e27991d596ab03558ebc406172f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:970cf89c89308db2915aa8883907d3e954d0efe023aa5924ef41f166140397cc
+size 107499
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..f60bfc16035c0d3dd14f21ea175d3b4ba32869c8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are standing outside a coffee shop. The man is wearing a black t-shirt and a baseball cap, while the woman is dressed in a white striped top. They are both looking at something off-camera, possibly engaged in a conversation. The coffee shop has a sign that reads \"FREE COFFEE COPS\", suggesting that they might be offering free coffee to law enforcement officers. The setting appears to be a sunny day, with the man and woman standing in the shade provided by the coffee shop. The overall style of the video is casual and relaxed, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "4F20DwUrZJ8_5_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman standing outside a building. The man is wearing a black t-shirt and a cap, and the woman is wearing a white striped shirt with her hair in a bun. They are positioned next to each other, fulfilling the 'Object(s): A man and a woman' condition. Additional elements like the sign and background are present but do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man and a woman standing side a building. The man is wearing a black t-shirt and a cap, while the woman is wearing a white striped top. sunglasses hair down a pony. The are positioned side to each other, which the 'Object(s)' A man and a woman.' condition.\"\n There elements like the building in the do present but do not conflict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2124074621491d6131a55787f685837348cb396e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d4ff2356b13198749e6d3b6ffb8b68fac330db72f0719eb6e3c1e974454cb32b
+size 186162
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..f3a187eb65ca85dcf87aa5e61968ba461c5ce36f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two young women are seen in a historical setting, possibly a village or town. They are dressed in period clothing, with hats and dresses that suggest a time in the past. The women are smiling and appear to be in a good mood, possibly enjoying a moment of leisure or celebration. In the background, there are other people, some of whom are also dressed in period clothing, and a horse is visible, further emphasizing the historical context of the scene. The overall style of the video is realistic, with attention to detail in the costumes and setting, suggesting a high-quality production."
+    ],
+    "video_ids": [
+        "JR10MgVJ580_16_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two young women in period clothing (hats and dresses), smiling and appearing in a good mood.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two young women dressed in period clothing, including hats and dresses, and they are smiling and appear to be in good spirits. Their expressions and body language convey happiness and engagement, matching the description. Additional elements like background people and a horse do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two individuals women wearing in period clothing, including hats and dresses, which they are smiling and appear to be in a spirits. The attire and body language suggest a and content, which the description provided The elements such the activity and buildings festive in not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..47f4b7cceab6218509e5e06413bb8921fe3621f8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f606a421515ac1be4e7eb5c67cd68bee0b4ce49b2879bdbd9b6a8c5681ec1f0d
+size 228239
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..1acfe9401bf0d10d57cb8d244465713a82c507f9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment on a soccer field. The main focus is a soccer player, dressed in a black and white striped jersey, who is in the midst of a powerful kick. His body is angled towards the right side of the frame, suggesting a strong, well-executed kick. The player's arms are outstretched, adding to the sense of motion and energy in the scene.  The background of the video reveals a large crowd of spectators, their faces a blur of anticipation and excitement. The crowd fills the stands, creating a sea of colors and shapes that contrast with the player's focused action.  The style of the video is dynamic and energetic, capturing the intensity of the game and the player's skill. The focus on the player's action, combined with the blurred background, draws the viewer's attention to the main event - the kick. The video does not contain any text or additional elements, keeping the focus solely on the action on the field."
+    ],
+    "video_ids": [
+        "5Q2aQgDWAUI_26_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A soccer player in a black and white striped jersey, a large crowd of spectators.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a black and white striped jersey (specifically, a Paris Saint-Germain home kit with 'Fly Emirates' sponsorship), and there is a large crowd of spectators in the background, which matches the 'Object(s)' condition. The presence of other players and a referee does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a soccer player wearing a black and white striped jersey,likelyally, a Juventus Saint-Germain jersey kit), thePS Emirates' sponsorship), which there is a large crowd of spectators in the background. which align the descriptionObject(s)' condition described The player of the elements or the soccer in not contradict the description description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ccc1229fcdbe300e5c9a6e3ff3aa8da8dfc09702
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e773e1f13d5537c10fd2425ab0a26913c9861baeb3f3ebd9568052ade21f9bbf
+size 237946
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..c92207edad88303639eb2b6bde943339556d1511
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a young man in a sports jersey, walking confidently towards the camera. He is waving his hand, possibly acknowledging the crowd or his teammates. The jersey is white with blue and black accents, and the word \"Dafabet\" is prominently displayed on the front. The man's expression is serious, suggesting he is focused on the game or event. The background is blurred, but it appears to be an outdoor sports stadium, indicating that the man is likely a professional athlete. The style of the video is dynamic and energetic, capturing the man in motion and conveying a sense of anticipation and excitement."
+    ],
+    "video_ids": [
+        "I7d-nzbw840_32_0to161"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Young man in a sports jersey.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man wearing a sports jersey with visible team branding and sponsor logos, consistent with the description. The setting appears to be a stadium, which aligns with the context of a sports jersey being worn. There are no elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man wearing a sports jersey with a branding logos and a logos. which with the description of The setting appears to be outdoors sports or which iss with the sports of a sports event. worn. The are no elements in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9b1bb2771f984b54ad43464f619f1eed8abd7bf0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1194-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee74d9123ab960a520fbe354887df4f8916fe4ba691ccd26a8fb8398d4da3fa7
+size 186156
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..9c0d6585ed820812b4decba3f4902290f4f52e2f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen in a bathroom, washing his hands at a sink. He is wearing a gray t-shirt with the word \"Guess\" written on it. The bathroom has a white sink and a silver faucet. On the sink, there are several bottles, possibly containing soap or lotion. The man is standing in front of a mirror, which reflects the image of the bathroom. The style of the video is a simple, everyday scene captured in a real-life setting."
+    ],
+    "video_ids": [
+        "D1MZjBp8L4U_2_114to271"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a gray t-shirt with 'Guess' written on it, a sink, several bottles, a silver faucet, and a mirror.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a gray t-shirt with 'Guess' written on it, standing at a sink. There is a silver faucet, a mirror, and several bottles (including soap and sanitizer) visible on the counter. All elements described in the condition are present and consistent with the video content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing a gray t-shirt with 'Guess' written on it. standing in a sink. There are a mirror faucet and and mirror above and several bottles onlikely what dispens possibly) on on the sink. The these match in the condition are present and match with the video content.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..92a99e55c2eae9046beaaac72939d76f81287a4e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e69e033aa038bc6c3dd6ad1db132a1de831765071df89163fd243a8dc41af71
+size 187783
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..6451a585646e5866bf874eba511050fe199e5c02
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment on a soccer field. A player, wearing a white jersey with the number 4, is the main focus. He is crouched down on the lush green field, his head bowed in disappointment. His teammates, also in white jerseys, are scattered around the field, some standing and others walking. The background is filled with spectators, their faces a blur of anticipation and excitement. The atmosphere is tense, the air thick with the anticipation of the next move in the game. The video is a snapshot of a moment in a soccer match, capturing the emotions of the players and the spectators alike."
+    ],
+    "video_ids": [
+        "kwDhlSQKuwc_17_0to154"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a white jersey with the number 4, crouched down with bowed head; teammates in white jerseys scattered around, some standing and others walking.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a white jersey with the number 4, crouched down with his head bowed, which matches the description. In the background, there are teammates in white jerseys scattered around, some standing and others walking, which also aligns with the description. The scene is consistent with the requested object conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a player in a white jersey with the number 4, crouched down with a head bowed, which matches the '. The the background, there are other in white jerseys, around, with standing and others walking, which also aligns with the description. The presence appears set with a provided conditions condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c4a99ceb4561f1d41199f69d415b517598af063a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c2660de56b42283dfcc147af39b45a9e097528802db5bbf6b638dbc266b471ae
+size 207449
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..b8af0f7f0a091c631bd6fc0c4cc7810af0763fb5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video opens with a serene beach scene, where the sandy dunes stretch out under a clear blue sky. In the foreground, several colorful plastic toys are scattered on the sand, including a bright red toy rabbit, a purple turtle, and a blue toy figure. An orange Hot Wheels track lies diagonally across the frame, leading towards the background. A small blue car is positioned at the end of the track, poised to roll down. As the video progresses, the camera remains stationary, capturing the tranquil beach setting. The text \"HOT WHEELS BEACH COASTER\" appears prominently over the scene, indicating the playful theme. The car then begins its journey down the track, rolling smoothly over the sand, creating a sense of motion and excitement against the calm backdrop of the beach."
+    ],
+    "video_ids": [
+        "410f80d55a48cc19e3056dc235a46889bb3fd2007fc69bcd25c1c0fc67b33b0a"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Several colorful plastic toys (bright red toy rabbit, purple turtle, blue toy figure), an orange Hot Wheels track, and a small blue car.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows several colorful plastic toys including a bright red toy rabbit, a purple turtle, and a blue toy figure, along with an orange Hot Wheels track and a small blue car. These elements are clearly visible and match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful several colorful plastic toys, a bright red toy rabbit, a purple turtle, and a blue toy figure. an with an orange Hot Wheels track and a small blue car. The elements match clearly visible and match the description provided.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..aeae37e785d69717c21406fc764b9513128560e0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3691c9f4cac49afd7fbe50e42b0c556ad9a1e410fa1c8611e83ee8fafee71ead
+size 219262
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..9586f1a86030e313bd300680a682544ac8009ebc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a bee's interaction with a yellow flower in a grassy field. The bee, with its black and yellow body, is seen hovering over the flower, landing on it, and then taking off. The flower, with its vibrant yellow petals, stands out against the green grass. The bee's movements are captured in detail, showcasing its agility and precision. The background is a blur of green, emphasizing the bee and the flower as the main subjects of the video. The overall style of the video is a close-up, slow-motion shot, allowing the viewer to appreciate the intricate details of the bee's actions and the beauty of the flower."
+    ],
+    "video_ids": [
+        "jUxfTbAU7rk_39_0to139"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bee and a yellow flower.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a bee hovering and landing on a yellow flower, which matches the 'Object(s)' condition. The background is a blurred green field, which does not contradict the description. The bee and yellow flower are the central focus and are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a bee and over landing on a yellow flower, which matches the descriptionObject(s)' condition. The bee is blurred blurred field field, which is not contradict the presence. The presence and the flower are the primary elements, fulfill present depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fe57ba66d8c043fb661fbc2db6b0404e08df22e7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ca33899337f345060ecf11030c12a29ed03fd82c7c04befbffee5e037d22f9d5
+size 131693
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..9706f0a4c336fdac8f49f943b1f3d1c173b5de6f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car from the perspective of the driver's seat. The car is a modern, dark-colored vehicle with a sleek design. The steering wheel is prominently displayed in the center of the frame, featuring a logo on the top. The dashboard is visible, with various controls and indicators. The car's door is open, revealing a window and a side mirror. The interior is well-lit, suggesting that the car is parked in a brightly lit area. The style of the video is straightforward and unembellished, focusing on the car's interior without any additional context or action."
+    ],
+    "video_ids": [
+        "AF1VbNDoKk4_24_0to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, dashboard, car door, window, side mirror? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the steering wheel, dashboard, car door, window, and side mirror as requested. All these elements are visible and accurately represented within the frame, fulfilling the 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows the interior wheel, dashboard, car door, window, and side mirror, described. The these elements are visible and match depicted in the frame, fulfilling the 'Object(s)' condition.\"\n any or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b0dae3a566bc355ce50033ed2dc24add58c0617e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba0b127a66824fae216d94a8c306ff154ce635d0395b8384b4857b6068b12156
+size 150228
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..8d36fe2970fdaaa196b397847726dd297669b64c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a CGI animation of a futuristic spacecraft in orbit around Earth. The spacecraft is cylindrical with a large central ring and multiple smaller rings around it. The exterior of the spacecraft is white with various windows and panels. The Earth is visible in the background, showing a blue sky with clouds and the curvature of the planet. The style of the video is realistic with attention to detail in the spacecraft's design and the depiction of Earth's atmosphere. The animation is smooth and the spacecraft appears to be in motion."
+    ],
+    "video_ids": [
+        "22TcDcy_qJs_10_0to152"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A cylindrical spacecraft with a large central ring and multiple smaller rings around it. The exterior is white with windows and panels. Earth is visible in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a cylindrical spacecraft with a large central ring and multiple smaller rings around it, matching the described structure. The exterior is predominantly white with visible windows and panels. Earth is clearly visible in the background, fulfilling the background condition. Additional elements like internal components and distant satellites do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a cylindrical spacecraft with a large central ring and multiple smaller rings around it. which the description structure. The exterior is predominantly white with visible windows and panels, The is clearly visible in the background, fulfilling the condition condition. The elements like the structures or a structures are not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6cf7c0d9db864aca537ece259b5d93dc7c34d33d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f4f670797514a7a3eb8f6fa54facb28b870e94799605484cf524ec9e21c5d44
+size 258029
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..b661c92b29ee3fd06bf7cfd7493f5e142941b603
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a pig walking through a muddy field. The pig is white with black spots and has a pink nose. The field is muddy and wet, with puddles scattered throughout. In the background, there is a barrel and a pile of rocks. The pig is walking towards the camera, and its head is turned to the side. The video is shot in a realistic style, capturing the pig in its natural environment."
+    ],
+    "video_ids": [
+        "6IFUxdeaGK0_48_0to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white pig with black spots and a pink nose.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a pig with a predominantly white body, black spots on its head and back, and a distinct pink nose, which matches the description. The pig is the central focus of the video, and no elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a white that a predominantly white body, black spots, its back, back, and a pink pink nose. which align the description provided The pig is walking central object of the video, and the other contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..774169183d744a14b87113cc267a2ea55df6f18f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7cb451071bc991dcf0de43a615235ab768ac53fcac2fa95e263a98dc9da05a0f
+size 249345
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..cd7107abb65d92fee535854f0997ee72ce63e9a4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a grey shirt and sunglasses standing next to an open car hood. He is gesturing towards the engine, possibly explaining something or pointing out a specific part. The car is parked on a gravel driveway, and the man appears to be in the middle of a conversation or demonstration. The style of the video is casual and informative, with a focus on the man and the car. The setting is outdoors, and the lighting suggests it might be daytime."
+    ],
+    "video_ids": [
+        "HqiZHAJtvWs_13_48to234"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a grey shirt and sunglasses, an open car hood, and a car parked on a gravel driveway.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a grey shirt and sunglasses, standing in front of an open car hood. The car is parked on a gravel driveway, as seen in the background. All core elements of the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two man in a grey shirt and sunglasses standing standing next front of a open car hood. The car appears parked, a surface driveway, which indicated in the background. The elements elements of the description are present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..db966bc3e457a4f2daf2240f73386c27bdd06554
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1195-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ca8af9c7a98631fb94b48c8698646d58aa43986796fdda7323afeb75a923f61d
+size 202598
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..bfed490734c29713b069171524643066120b4746
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man with a beard and sunglasses is seen holding a glass of whiskey. He is wearing a patterned suit and is holding a microphone in his other hand. The man is standing in front of a blue background and appears to be speaking or singing. The video captures the man's confident and relaxed demeanor as he enjoys his drink and engages with the audience. The overall style of the video is casual and relaxed, with a focus on the man's performance and the atmosphere of the event."
+    ],
+    "video_ids": [
+        "Hdx3tZtwiNs_10_17to150"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and sunglasses, a glass of whiskey, a patterned suit, and a microphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and sunglasses, holding a glass of whiskey, wearing a patterned suit, and standing in front of a microphone. All core elements described are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man with a beard and sunglasses, holding a glass of what and and a patterned suit, and speaking with front of a microphone. The the elements of in present in match depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..492d1eef848cadad94409cd60262ec8c18294878
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7cc40bd8dcbc6271b9f480f5b3a7d9f61c93f0f1cbff42f202a4277c4d457e4c
+size 467055
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..699cf7adc116ee0a026a8788a7cee08c8d2a534c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man standing in a grassy field with trees in the background. He is wearing a gray t-shirt and has a beard. In the first frame, he is looking directly at the camera with his arm down. In the second frame, he is speaking and his arm is slightly raised. In the third frame, he has his arm fully extended towards the camera, gesturing as he speaks. The style of the video is casual and informal, with the man appearing to be engaged in a conversation or presentation. The setting is outdoors and the lighting is natural, suggesting that the video was taken during the day."
+    ],
+    "video_ids": [
+        "1jB16_IKOGY_38_676to866"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man as the central subject, speaking to the camera and gesturing with his hand. The description of 'A man' is accurately fulfilled, as there are no contradictions or misrepresentations of the main object in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a man standing the main figure. standing and the camera in makinguring with his hands. The presence of theA man' is fulfilled represented by as the is no additional or additionalrepresentations in the subject subject in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7a17d570bc7434ccecbe9212a24a17998b3bc636
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67240095fd022bbffcb455d65ff518ef006190ab40a3afeb3a10d9d653462808
+size 176935
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..085e5a56a9a1e2007c40e3e6595356b25a09069d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a female character in a stylized, cartoon-like environment. She is wearing a white hoodie with a red and yellow design, resembling a fast-food logo, and a yellow beanie with googly eyes. The character has brown hair and is carrying a blue backpack. She is standing in front of a building with an orange and white facade. The overall style of the video is vibrant and colorful, with a playful and whimsical atmosphere."
+    ],
+    "video_ids": [
+        "X2WTwrFf2Q4_30_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A female character wearing a white hoodie with a red and yellow fast-food logo design, a yellow beanie with googly eyes, and a blue backpack.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a female character wearing a white hoodie with fast-food-themed logos (burgers and fries), a yellow hood with googly eyes, and a blue backpack. These elements match the description provided, even though the hoodie's design includes multiple colorful food items rather than a single red and yellow logo. The core elements of the description are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a female character wearing a white hoodie with a-food logo logos (reders and fries), a yellow be with googly eyes, and a blue backpack. The elements match the description provided, fulfilling though the hoodie's design is additional logos logos items, than just single red and yellow logo. The presence elements of the description are accurately represented.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..980147cb96be6d7f99f95d0c7ad36b13792e563e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe550544188ff3eb04f403a58c8827a24e1abf12a032d39a152ad53317874f55
+size 127283
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..bdef98a866d2ab6e0a02afee38aafa29b013b258
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a promotional advertisement for a car, showcasing its performance capabilities. The car is a Mercedes-Benz, and it is driving down a city street with tall buildings in the background. The car is in motion, and the camera follows it from a distance, capturing the speed and power of the vehicle. The car's performance statistics are displayed on the screen, indicating its horsepower and torque. The city street is lined with palm trees, adding a tropical feel to the scene. The overall style of the video is sleek and modern, with a focus on the car's performance and the city's architecture."
+    ],
+    "video_ids": [
+        "7loYredzxQE_46_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Mercedes-Benz car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a Mercedes-Benz car driving down a city street, which matches the 'Object(s)' condition. The car is clearly visible and central to the video's content, with no contradictions to the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a Mercedes-Benz car, through a city street. which directly the 'Object(s)' condition. The car is clearly visible and identifiable to the scene, content, with no conflicting to the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4e6c63f948ef70dfefa68eea117c7207f4ea6e98
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a52aea7dd5d4b896217c01612bcd305cf9ca9a62aecdbe3da47d475c044c806e
+size 353807
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..5f8ec867aa7d9c7983acd9c8769e852e96f5b0b9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features two animated characters, both wearing purple hooded cloaks and glasses. They are standing on a wooden platform with train tracks in the background. The character on the left is making a surprised face, while the one on the right is gesturing with her hands. The scene appears to be set in a forested area with trees in the background. The characters are likely engaged in a conversation or reacting to something happening off-screen. The style of the animation is colorful and cartoonish, with a focus on exaggerated facial expressions and body language to convey emotion."
+    ],
+    "video_ids": [
+        "6F1C1JEzlq8_248_0to186"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two animated characters wearing purple hooded cloaks and glasses.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features two animated characters wearing purple hooded cloaks, as described. One character also wears glasses, matching the condition. Although the scene briefly shifts to a bear sleeping, the primary focus on the two characters in purple hooded cloaks and glasses fulfills the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two animated characters wearing purple hooded cloaks and which described. Both character has has glasses, which the description. The the video includes shows to show different,, which focus focus remains the characters characters in purple cloed cloaks and glasses confirmsfills the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c0985c886216cbf4652a8b3a3a7b753b7019795b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0114e30efbade88bf85fa3697e2e6d6cd28cd600b5ea6b067cf978f109de58e2
+size 209439
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..29810060885d4558a6f8f3d56fc7e90d27770f9d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a dynamic and stylish showcase of a red sports car in motion. The car, with its sleek design and shiny exterior, is the main focus of the video. It is captured in three different frames, each highlighting a different aspect of the car's design and movement.  In the first frame, the car is seen from a side angle, showcasing its aerodynamic shape and the smooth lines that run along its body. The car's red color stands out against the backdrop of a building, drawing attention to its design.  The second frame captures the car from a front angle, emphasizing its powerful stance and the sharp angles of its headlights. The car's black rims add a touch of contrast to its red exterior.  The third frame is a close-up of the car's front, highlighting the intricate details of its design, such as the grille and the headlights. The car's shiny exterior reflects the surroundings, adding depth to the image.  Throughout the video, the car is seen in motion, driving on a road, which adds a sense of speed and excitement to the video. The car's design and the way it moves are the main elements of the video, creating a dynamic and stylish showcase of the car's features."
+    ],
+    "video_ids": [
+        "aXug1j2kQPg_16_0to157"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by prominently featuring a red sports car. The car is clearly visible and occupies the central focus of the frames, matching the description. Additional elements like the background building and greenery do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features fulfills the 'Object(s)' condition as depicting featuring a red sports car. The car is shown visible in occupies a majority focus of the frames, with the description of The elements such the building and and theery do not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4b8fb8cd50a467081f9b5031edfb91cf6ee59896
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e856a5ef060276c4b963fad82036dc8637593e8f507f2d645439c76083ecb2e9
+size 225656
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..7d814695da35d33433a6d100218a5d0dd0e3505b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a vibrant underwater scene featuring a blue and yellow striped fish swimming amidst a coral reef. The fish, with its striking colors, is the main focus of the video, swimming gracefully through the water. The coral reef, with its intricate patterns and shapes, provides a beautiful backdrop to the fish's journey. The water around the fish is clear, allowing for a full view of the fish and the surrounding environment. The video is shot from a perspective that places the viewer right in the water with the fish, providing a sense of being right there in the underwater world. The overall style of the video is one of natural beauty and tranquility, capturing the essence of life beneath the waves."
+    ],
+    "video_ids": [
+        "Lxi6ZdgqhOU_32_104to273"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Blue and yellow striped fish, coral reef? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows blue and yellow striped fish (specifically, blue tangs with yellow stripes) swimming near a coral reef, which matches the 'Object(s)' condition. Additional elements like other fish and water surface ripples do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a and yellow striped fish swimminglikelyally a a tangs) yellow stripes) swimming among a coral reef. which matches the descriptionObject(s)' condition. The elements like the marine or marine details detailsipples are not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ad6f4e2449a3c408c0eb8272d30d18d0b22bcb29
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:61f015d5e5080260157fad5d8a17ea37b743bd73421bb365523dec2dc19a8a32
+size 130811
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..8e45d8489bbcb113cc1702b16c86b73abcc00414
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man fishing in a river. He is wearing glasses and a white shirt with a green patch on the sleeve. He is holding a fishing rod and a fishing net. The river is surrounded by rocks and trees. The man is standing on the bank of the river. The video is shot in a realistic style, capturing the man's actions and the natural environment."
+    ],
+    "video_ids": [
+        "lXMhmU3MX4g_49_16to151"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a fishing rod, a fishing net, glasses, a white shirt with a green patch on the sleeve? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing glasses and a white shirt with a green patch on the sleeve, holding a fishing rod and with a fishing net attached to his back. All core elements described are present and consistent with the video content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man fishing glasses and a white shirt with a green patch on the sleeve, holding a fishing rod. a a fishing net attached to it rod. The the elements of in present in accurately with the video content.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..50a0b6f571b65aab5702f6a8786b6c95ca2fd437
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1196-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:56248ab8257201cf2a4c5be642877eaa4224e0e833b408908eb3d74d2d657c72
+size 248966
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..b109c8e1b7b8b8fd37135ed1b2a22b63e9c32605
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up of a tiger in a zoo enclosure. The tiger is seen in three different frames, each showing a different expression or action. In the first frame, the tiger is seen with its mouth closed, looking directly at the camera. In the second frame, the tiger's mouth is slightly open, and it appears to be growling. In the third frame, the tiger's mouth is wide open, and it seems to be roaring. The tiger's fur is a mix of black and brown stripes, and its eyes are focused and intense. The background of the video is blurred, but it appears to be a zoo enclosure with trees and foliage. The style of the video is realistic and captures the tiger's natural behavior in a controlled environment."
+    ],
+    "video_ids": [
+        "2c44jJfDVl0_51_0to122"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A tiger? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a tiger with distinct orange fur, black stripes, and facial markings, which matches the description of 'A tiger'. The tiger is the central and only prominent object in the video, and there are no elements contradicting this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a tiger-up of a tiger's a features fur, black stripes, and white features. which are the description of aA tiger'. The tiger's shown central focus only object object in the video, and there are no additional thating the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c6a8d64e3c5f0f38fd6b0df26a72ca2971dee785
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26120574dbe1d60bc8de2dae840f72b51cbadd378bb5595e759c5f6842250673
+size 330555
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..cb9e64ac49239edabe12c12bf58a433768072cab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of triumph for a soccer player. The player, dressed in a green and white striped jersey, stands on the field with his arms outstretched in a victorious pose. The background reveals a stadium filled with spectators, their faces a blur of anticipation and excitement. The player's position on the field, coupled with his triumphant pose, suggests that he has just scored a goal or achieved a significant victory in the game. The overall atmosphere of the video is one of celebration and achievement, encapsulating the thrill and excitement of the sport."
+    ],
+    "video_ids": [
+        "0dSbQr3t3k4_24_0to129"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A soccer player in a green and white striped jersey with arms outstretched.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a green and white striped jersey with arms outstretched, which matches the core description. The background, including spectators and stadium elements, does not contradict this description and is consistent with a soccer match setting.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a green and white striped jersey, arms outstretched. which matches the description description. The player, though the and the elements, does not contradict the description.\"\n is acceptable with a typical match setting.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..54f6557418558d1f93457380b9a0738a022b8d23
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5134a82cad6630b84b1968467865965b8c0cec4f0041ce879c0edbcc3c747957
+size 205529
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..adf6a81897cc86f6773c6a547231784824e4cd7c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the process of pouring a rich, brown curry from a metal pot into a bowl. The curry is thick and appears to be a type of Indian dish, possibly Dal makhani, as indicated by the text overlay. The pot is placed on a metal tray, and the curry is being poured into a bowl that is positioned to the right of the pot. The background is blurred, but it seems to be a kitchen setting with other pots and pans visible. The style of the video is simple and straightforward, focusing on the action of pouring the curry without any additional context or embellishments. The lighting is bright, highlighting the texture and color of the curry."
+    ],
+    "video_ids": [
+        "VfNqA2ukNrk_182_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A metal pot containing curry and a bowl positioned to its right.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a metal pot containing dal makhani curry, and to its right, there is a bowl with sliced onions, which matches the described objects. The presence of other elements like additional bowls and utensils does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a metal pot containing a (akhani, being which a the right, there is a bowl positioned the onions. which align the description setup. The text of the elements like the text or textils does not contradict the core description.\"\n}<|im_end|>\nGuidId",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..523bdecf46bac2de6ba3d79187b39e476ec4c1d4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b81c21952a192254d9862db200d03ce823bec904d3a3f1af47117b550226573d
+size 183869
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..e76da84ffac6f1e93196e75c1891b64e29bfdd9b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two young children are seated at a dining table, enjoying a meal of cereal. The boy, dressed in a yellow shirt, is on the left side of the frame, while the girl, wearing a pink shirt, is on the right. They are both using spoons to eat their cereal, which is served in white bowls. The table is set with a white tablecloth, and there are two bottles of milk on the table, one near each child. The children are seated on gray chairs, and the scene takes place in a room with a white wall in the background. The video captures a simple, everyday moment of childhood, with the children enjoying their meal together."
+    ],
+    "video_ids": [
+        "35CYrD03DJo_38_0to189"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two children, a boy in a yellow shirt and a girl in a pink shirt, two spoons, two white bowls of cereal, two bottles of milk, and two gray chairs.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two children, a boy in a yellow shirt and a girl in a pink shirt (the girl's shirt has a pink 'Barbie' design), sitting at a table with two white bowls of cereal, two spoons, and two bottles of milk. They are seated on two gray chairs. All specified objects are present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two children, a boy in a yellow shirt and a girl in a pink shirt,matching ' is shirt is a pink andHbie' logo), sitting at a table. two white bowls of cereal, two bottlesons, and two bottles of milk. There are also on gray gray chairs. The the objects are present and correctly the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..19e5ad68574295d37dc43b192d6f1f6f711ed449
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:15236ce45a545d94714f9326516ac2d5b6b1fc1285309ad1c2522bfbf30c67fb
+size 105712
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..7b6762b4483a67a68359a88f425aeaa2ac91893d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a chocolate dessert being cut with a knife. The dessert is placed on a white plate, and the knife is being used to slice through the chocolate. The chocolate has a glossy finish and is adorned with chocolate chips. The knife is silver and has a pointed tip. The background is blurred, but it appears to be a kitchen setting. The style of the video is a slow-motion shot, capturing the details of the dessert and the knife as it slices through the chocolate."
+    ],
+    "video_ids": [
+        "Iv4mQJsBTPk_331_775to958"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chocolate dessert, knife, chocolate chips? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chocolate dessert in the form of a high-heeled shoe, which matches the 'Chocolate dessert' condition. There are no knives visible in the video, but since the presence of additional elements is acceptable as long as they don't conflict with the core description, this is not a contradiction. Chocolate chips are not explicitly visible, but the small chocolate droplets being dispensed can be considered a form of chocolate chips or similar chocolate elements. Thus, the core description is largely fulfilled.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chocolate dessert being a process of a cake-geled shoe, which is the 'Chocolate dessert' condition. A is chocolate additional or in the video, but the the question of a elements ( acceptable, long as they do't conflict with the core description, the can not a contradiction. The chips are not present mentioned, but the dessert chocolate piecesplets on cutensed from be interpreted as form of chocolate chips, chocolate decorative decorations, Therefore, the video elements is largely matched.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..378aa7dc27ccd989c2d38bafc83ee140dfa7c57a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c4751b861ce66e724a2f9e7b03f3d33a91356698739af07c0ccf2e7afe51ae3b
+size 114047
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..f15596204d88001ae2956c59d8f46c3fd6ab391b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman posing against a gray wall. She is wearing a black hoodie and pants, with her hood up. Her hands are placed on her head, and she is looking down. The style of the video is casual and relaxed, with a focus on the woman's outfit and pose. The gray wall provides a neutral background that allows the viewer to focus on the woman and her attire. The overall mood of the video is laid-back and comfortable."
+    ],
+    "video_ids": [
+        "HiHsnQt4uHg_9_24to155"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Woman posing with black hoodie and pants, hood up, hands on head, looking down? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a black hoodie and pants, with the hood up, her hands on her head, and she is looking down. These elements match the description exactly, and there are no conflicting elements in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a black hoodie and pants, with the hood up, and hands on her head, and she is looking down. The elements match the description provided, indicating there are no additional elements in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..710c723c33248cba0fca6f3cd9cc9b23c5e680e5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe6c2adb7c82fe58d7f81bb8c51069dd4b8e762a67d4bffc806ea21f4804f119
+size 56864
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..8ff273d9699c81b14a2cce7864a9569d2f59b972
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a smartphone lying on a white surface. The phone has a black back with a colorful floral design. The camera is positioned to capture the phone from a top-down perspective, focusing on the back of the device. The phone is connected to a charger, which is plugged into the phone's charging port. The charger has a USB-C connector and is connected to the phone via a cable. The phone is in the process of charging, as indicated by the illuminated battery icon on the screen. The video is a simple product demonstration, showcasing the phone's design and charging capabilities."
+    ],
+    "video_ids": [
+        "tmb03ujYI04_81_0to152"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Smartphone, charger, cable? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a smartphone with a visible charging port and speaker grilles, which aligns with the 'Smartphone, charger, cable' condition. Although no actual charger or cable is shown, the charging port implies the presence of a charger and cable, fulfilling the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a smartphone with a colorful screen cable, a,ills, which ares with the 'Smartphone' charger, cable' condition. The the charger charger or cable is shown, the presence port and the presence of a charger, cable, which the '.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5599bfcb9d76dc863dcfdacc041e69602a55f486
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:91ba2ac3c1f4634588dbcdf7f76145fd8f343b86d75e358c6eb14777d95903c7
+size 170025
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..cfb3e11a201e4cc9a05419fef1d7c3df2b2a7b16
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a plate of French toast with whipped cream and orange zest. The French toast appears to be golden brown and crispy on the outside, while the inside is soft and fluffy. The whipped cream is generously dolloped on top of the French toast, and the orange zest adds a pop of color and a hint of citrus flavor. The background is blurred, but it seems to be a kitchen setting with a bottle of syrup on the counter. The style of the video is a food close-up, focusing on the textures and colors of the dish. The lighting is bright and even, highlighting the details of the French toast and the whipped cream. The video does not contain any text or narration."
+    ],
+    "video_ids": [
+        "wjc-z3mG3nA_28_0to170"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: French toast, whipped cream, orange zest? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows French toast rolls topped with whipped cream and garnished with orange zest. The presence of a Grand Marnier bottle and oranges in the background supports the orange flavor theme without contradicting the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a toast topped, with whipped cream and garnished with orange zest. The presence of a jar Marnier bottle in a in the background suggests the idea zest theme, contradicting the core description.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..14292975106980e8d0aa48993824e71ba8fde65c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1197-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5624385c933e9fd89494775031af2f5e944ced1a9a4fefad379322b26ff80564
+size 105573
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..cdb6835b790e723891166817cd31eac1d4c8c72b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young woman with long brown hair is seen in a kitchen, preparing a meal. She is wearing a white shirt and is standing in front of a wooden counter. On the counter, there are several bowls and bottles, as well as a blender. The woman is holding a spoon and appears to be stirring something in one of the bowls. The kitchen is well-lit, with natural light coming in from a window. The overall style of the video is casual and homey, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "9jKtf27L2u8_9_90to254"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman, several bowls, bottles, a blender, a spoon? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a young woman standing in a kitchen setting. Several bowls (metal and white) are visible on the counter in front of her. Bottles are present in a rack on the counter behind her. A blender is also visible on the counter. Although a spoon is not clearly visible in the provided frames, the presence of other specified objects confirms the video largely matches the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a young woman in at a kitchen,. She bowls arebothlic ceramic) are visible on the counter, front of her. Thereles are also, the blender in the counter behind her. A blender is also visible on the counter to The a spoon is not explicitly visible, the image frames, the woman of the kitchen objects ( that video largely matches the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..18987c04731ca5d9dde661a929ddf9b014c8796c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60c9bab679a41059334c27badcca8c42e6ee680b372d8e44d21869383d102f28
+size 110390
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..4550d8e55bcf18e67145c7a681c847109dc12894
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man giving a presentation on a stage. He is dressed in a black suit and a purple shirt, and he is wearing glasses. He is holding a cell phone in his hands. The stage has a red curtain in the background. The man appears to be speaking to an audience, and he is gesturing with the cell phone as he speaks. The style of the video is a straightforward presentation, with the man as the main focus. The lighting is bright, and the stage is well-lit, highlighting the man and the red curtain in the background. The video does not contain any other objects or people. The man's speech is the main focus of the video."
+    ],
+    "video_ids": [
+        "WbCguICyfTA_27_0to135"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black suit and purple shirt, holding a cell phone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black suit and purple shirt, and he is holding a small object that appears to be a cell phone or remote. The description is largely accurate, with no significant contradictions. The presence of a microphone and stage setting does not conflict with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black suit and a shirt, holding he is holding a cell object that appears to be a cell phone. a. The setting is largely matched, and the significant contradictions or The man of a red stand the-like in not conflict with the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b05a503c427732c49c86cda76b47d285737cdc42
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:881699450e6e57a749d3fa2c136d6422d5af639dfeafba977de5e17e3368973d
+size 132750
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..ecbcd13d8036c77f71ee951ff73365a9b10e2176
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a scene of a yellow school bus parked in a lot. The bus is large and yellow, with a flat roof and a large windshield. It is parked next to a building, which is a large yellow building with a flat roof. The building has a large window and a door. The bus is parked on a concrete lot, which is surrounded by a fence. The lot is empty, except for the bus. The bus is parked in front of the building, and the building is behind the bus. The bus is the main focus of the video, and it is parked in the middle of the lot. The building is in the background, and it is behind the bus. The bus is parked on the concrete lot, and the lot is empty. The bus is parked in front of the building, and the building is behind the bus. The bus is the main focus of the video, and it is parked in the middle of the lot. The building is in the background, and it is behind the bus. The bus is parked on the concrete lot, and the lot is empty. The bus is parked in front of the building, and the building is behind the bus. The bus is the main focus of the video, and it is parked in the middle of the lot. The building is in the background, and it is behind the bus. The bus is parked on the concrete lot, and the lot is empty. The bus is parked in front of the building"
+    ],
+    "video_ids": [
+        "8H_RnDfO4t8_55_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Yellow school bus, large yellow building? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a yellow school bus parked in an outdoor area, and a large yellow building is prominently visible in the background. These core elements match the description, and while there are other objects present (like people, RVs, and rocks), they do not contradict the specified objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a yellow school bus parked in front area setting, which there large yellow building in visible visible in the background. The elements elements match the description provided and there there are additional elements like,such the and grasss, and a), they do not contradict the main '.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..610db40bfd472c7a0388aa4fbc46a0be09ef7e7b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e92066833bbf4318740b53bba94fcc59c040189d00d79bb41e41a9c29bedac2
+size 167282
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..aa9b472a854fb42aa8c91d070431795c62ed7597
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a day in the life of a farmer, working diligently in a vast field. The farmer, driving a red tractor, is seen plowing the field, creating a neat pattern of furrows. The field is a vibrant green, indicating healthy crops. The tractor leaves behind a trail of freshly turned soil, adding a sense of motion to the scene. The farmer's journey takes him from the left side of the frame towards the right, suggesting a day of hard work ahead. The overall style of the video is realistic, capturing the essence of rural life and the dedication of farmers."
+    ],
+    "video_ids": [
+        "wayce268EGs_70_0to145"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Farmer, red tractor? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red tractor actively moving through a field, which matches the 'Object(s)' condition. Although the farmer is not clearly visible inside the tractor, the presence of the red tractor fulfills the core requirement. Additional elements like wind turbines and different crop fields do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red tractor driving working through a field, which align the descriptionFar(s)' condition of The the video is not directly visible, the tractor, the presence of the tractor tractor infills the condition description of The elements like the patterns and the colored patterns do not contradict the main but}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f530ae52ba2658b3da81a6cb44cba8361e9e7cb5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:74b03e28159f23cd4ed9b2e7b5fe0bf01d329aac9ea71bf674a39530cb93f8bf
+size 163197
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..d83eb4ce397e746061925583d8f5f542bc2bc7d0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a tranquil forest trail, with a wooden fence on the left side. The trail is surrounded by lush greenery and towering trees, creating a serene and peaceful atmosphere. The sun casts a warm glow on the path, highlighting the texture of the wooden fence and the softness of the grass. The trail leads deeper into the forest, inviting viewers to follow it and explore the natural beauty of the woods. The overall style of the video is calm and serene, capturing the essence of a peaceful forest walk."
+    ],
+    "video_ids": [
+        "qX4sHLNy7Dw_13_0to101"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Wooden fence, lush greenery, towering trees, grass? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a wooden fence running along a dirt path, surrounded by lush greenery and towering trees. Grass is visible on the slopes and ground, matching the described elements. No conflicting elements are present that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a wooden fence on along the path path, surrounded by lush greenery and towering trees. The is also on the ground and along, complement the description elements. The additional elements are present, would the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..183592ffc19a6104446866ef0ee482f974e8cbff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e11941e94ef01d376f9be620677848f284c70d2c2ee2833f0eb6f0ff12a5bd6
+size 278251
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..720978cdda64e5a81ac5fc64b986545b9530aed7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young man with a friendly smile, standing in front of a microphone. He is wearing a black t-shirt and appears to be in a room with a bookshelf in the background. The room has a warm and inviting atmosphere, with the bookshelf filled with various books. The man seems to be in the middle of a conversation or recording, as he is looking directly at the camera. The overall style of the video is casual and approachable, with the man appearing relaxed and comfortable in his environment."
+    ],
+    "video_ids": [
+        "VYSdrL4RtbE_43_0to154"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man with a friendly smile, a microphone, a black t-shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man with a friendly smile, wearing a black t-shirt with a graphic design, and he is positioned in front of a microphone. The core elements described are clearly present and consistent throughout the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young man with a friendly smile, wearing a black t-shirt, a microphone design, and speaking is positioned in front of a microphone. The background elements of in present present, match with the frames frames}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..117e3424a53f62af84ac8088c24a96d19b71c98a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7089262edccd4ea1e531a650f120228002a9dbbfd44569f6dfe577b3b8b0abe4
+size 94142
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..91244a7b6c872235d67c4f35e23cf4a2baeb9df8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a promotional advertisement for a blue BMW SUV. The car is shown driving down a dirt road in a rural area with tall grass and a clear sky. The car is sleek and modern, with a large grille and distinctive headlights. The car is driving towards the camera, giving a sense of motion and progress. The background features a city skyline in the distance, suggesting that the car is capable of handling both urban and rural environments. The video is likely designed to showcase the car's off-road capabilities and its stylish design."
+    ],
+    "video_ids": [
+        "LtXvZgnorTs_13_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue BMW SUV? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a blue BMW SUV parked on a dirt road, with the distinctive BMW grille and logo visible. The vehicle's color and model are consistent with the description, and no elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a blue BMW SUV, on a sandy path. which a iconic BMW grille and logo visible. The vehicle is color and model are consistent with the description of and there additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..174055caf265922f7e111f62b99fde4597ad6d0a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d05f816968de0c6cd41cc688667a3c9a867c8641bbf8939657313c98d5fd54c9
+size 374207
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..cf51c1d30fad3a09e2516124c3950ff1770ccc52
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two men are engaged in a conversation at a wooden picnic table. The man on the left, wearing a gray sweater, is gesturing with his hands as he speaks. The man on the right, wearing a red apron, is attentively listening. The table is set with plates of food, including a blue plate with a sandwich and a bowl of beans. There are also utensils such as a knife and a fork, as well as a cup and a bottle. In the background, there is a barbecue grill and a tree. The setting appears to be outdoors, possibly in a park or a backyard. The overall style of the video is casual and relaxed, capturing a moment of conversation between two friends over a meal."
+    ],
+    "video_ids": [
+        "ZeBJe92A0_c_21_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, a wooden picnic table, a blue plate with a sandwich, a bowl of beans, utensils (knife and fork), a cup, a bottle, a gray sweater, a red apron? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men standing at a wooden picnic table. There are blue plates with food (appearing to be sandwiches or grilled items), a bowl of beans, utensils (knife and fork), a cup, and a bottle (likely BBQ sauce) on the table. One man is wearing a gray sweater, and the other is wearing a red apron with 'BBQ Bootcamp' written on it. All specified objects are present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts two men sitting at a wooden picnic table. On is blue plates with a,aars to be a) similar items) a bowl of beans, utensils (knife and fork), a cup, and a bottle onpossibly a sauce or on the table. One man is wearing a gray sweater, and the other is wearing a red apron. aBBQ'y' written on it. The the objects are present and match the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6ffc1377ce4975fc51e4f98a4e2c5c1ad5835232
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1198-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3112dc050873aa26f9cb4c327ad0f8d78a5f7d9bdf84c1e0017b2127370d6362
+size 187432
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..55cbe6e46c836e9fe3ca5caf882f57e708fff7a4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a white dog interacting with a rectangular mirror placed on a concrete surface. The dog, with its fluffy fur, approaches the mirror and begins to walk along it, its reflection clearly visible. As the dog moves, its reflection shifts accordingly, creating an illusion of movement across the mirror's surface. The background features a mix of shadows cast by nearby objects, possibly furniture or trees, adding depth to the scene. The lighting suggests it is either early morning or late afternoon, as indicated by the long shadows and warm tones. The dog continues to explore the mirror, occasionally stopping to sniff or look at its reflection, showcasing a curious and playful demeanor. The overall atmosphere is calm and serene, with the dog's interaction with the mirror being the focal point of the video."
+    ],
+    "video_ids": [
+        "62bb4c2ed470fe643add29fb2cd846c1133f389e4aaf997709c87f0d16690b2c"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white dog and a rectangular mirror? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white dog interacting with a rectangular mirror placed on a concrete surface. The dog is clearly visible, and the mirror reflects its image, fulfilling the described condition. Additional elements like shadows and a wooden structure in the background do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white dog interacting with a rectangular mirror, on a tiled surface. The dog's seen visible, and the mirror reflects its image, fulfilling the ' ' of There elements like the and furniture bench bench in the background do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3bd611788da2828ae47bd8c3630f99eecc20aed7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:728a3134ac1f389c00dee1fe23fb8dce0efb0afe2cb5080781540972ced6282e
+size 146130
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..51eb9baf8a0af668b87ff1cbb387e8ed8764b644
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man stands confidently in a modern, high-rise building with a panoramic view of a city skyline. He is dressed in a green and black suit, complete with a cape and a scepter, giving him an air of authority and power. His long hair is neatly styled, and he gazes out over the city with a serious expression. The room he stands in is sleek and modern, with large windows that offer a stunning view of the city below. The man's position in the room and his attire suggest that he is a powerful figure, perhaps a leader or a superhero. The video captures a moment of quiet intensity, as the man contemplates the city below and the challenges that lie ahead."
+    ],
+    "video_ids": [
+        "VFG6ox59coE_45_0to102"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a green and black suit with a cape and scepter, long neatly styled hair, and a serious expression.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed in a green and black suit with a cape, holding a scepter, and has long, neatly styled hair. His expression appears serious, matching the description. The background and additional elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man wearing in a green and black suit with a cape, which a scepter, which has long, neatly styled hair. His expression appears serious. which the description. The setting, setting elements, not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9492d47afc5140f8e3f8d9c128ff5f74a6b6083c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:abfea75d21a8aaab816fc051df11d346dc931742d8cf74b836d82e32907d553f
+size 101910
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..3ed0451f56953374c3cc786442a1c00a075cf90e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is a quarterback from the New England Patriots, who is in the process of throwing the football. He is wearing a blue and silver uniform, complete with a helmet and a face mask. The quarterback is in the center of the frame, with his arm extended, ready to make the throw. The football is clearly visible in his hand, emphasizing the action of the throw. The background is filled with other players, all in various positions on the field, adding to the sense of movement and action. The style of the video is realistic, capturing the intensity and excitement of the game."
+    ],
+    "video_ids": [
+        "NfgcGLunymw_3_461to610"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A quarterback from the New England Patriots, a football, and other players on the field.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a quarterback wearing a New England Patriots uniform (number 12), holding a football, and other players in similar uniforms on the field. The visual elements align with the described 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a player in a blue England Patriots uniform,blue 13), holding a football, and there players on the uniforms on the field. The presence elements match with the description 'Object(s)' condition, contradict.\"\n}<|im_end|>\n addCriterion",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..69b70894f052260dda0e42499c51dc81c2448db0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cacf6b04061498e28c8992079fe5ca1977ae45b89b6a0c29f127f719d61f9073
+size 238351
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..0b17586525efd899cd95523d3b2d6964c68a239c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated sequence featuring a whimsical scene with a rocket ship, a barn, and a pile of hay. The rocket ship, painted in white and gray, is situated in the middle of the scene, with a large cloud of smoke billowing out from its base, suggesting it has just landed or is about to take off. To the right of the rocket ship, there's a barn with a blue roof, and a red wagon is parked in front of it. On the left side of the image, there's a large pile of hay, and a large yellow object resembling a sunflower is placed next to it. The background is filled with green trees and a clear blue sky dotted with fluffy white clouds. The overall style of the video is playful and cartoonish, with vibrant colors and exaggerated proportions."
+    ],
+    "video_ids": [
+        "I3syX-EBznI_54_0to113"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Rocket ship, barn, pile of hay, red wagon, large yellow object (sunflower-like).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a rocket ship, a barn, a pile of hay, and a red wagon as described. The large yellow object is not a sunflower but appears to be a pile of hay, which is consistent with the description. The presence of additional elements like daisies and trees does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a rocket ship, a barn, a pile of hay, a a red wagon. described. Additionally large yellow object, also explicitly sunflower but appears to be a sun of hay, which is also with the description. The presence of these elements like theandelies and a does not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6cb40f8528f610c925fe19e9ae9750fc9336c879
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1f0fc4db2bc1b7cba3791eda57222ae9ec7e9c490f4f6f8fa6de0aa3a42cc2aa
+size 181795
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..e5a96d6965fb0079158ae118d5f53eb9cb74e58f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a pair of brown leather shoes with a unique design, featuring a strap across the top and a buckle. The shoes are displayed on a gray surface, possibly a table or a stand. The style of the video is a close-up product shot, focusing on the details of the shoes, such as the stitching, the texture of the leather, and the buckle. The lighting is soft and even, highlighting the shoes without creating harsh shadows. The background is neutral and does not distract from the shoes. The video does not contain any text or additional elements. The overall impression is that of a high-quality, well-crafted product, likely intended for a luxury market or for showcasing the craftsmanship of the shoe maker."
+    ],
+    "video_ids": [
+        "qD7oM088gRg_13_103to239"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A pair of brown leather shoes with a strap and buckle.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a pair of brown leather shoes with a decorative strap and buckle, matching the description. The shoes feature a fringed strap with a gold buckle, which aligns with the 'strap and buckle' requirement. The presence of a wooden shoe tree inside one shoe does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a pair of brown leather shoes with a strap strap and a, which the description provided The shoes are a glossyinged design and a buckle buckle, and iss with the 'strap and buckle' part. The material of a signature sole tree in the of is not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6c57b9cde2db1e05a16b4b93ab3464107c8930a5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:edb5b8962e22ad72297270b095e2c60d27e55d1f6efa9810ae36a5c2a894c15d
+size 59788
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..a67d5219a9ec649aed9aee4479187b6d90fdd6ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a black Porsche sports car parked on a concrete driveway. The car is sleek and shiny, reflecting the surrounding environment. The car is positioned in front of a wooden fence, which adds a rustic touch to the scene. The car's design is modern and sleek, with a low profile and a streamlined body. The wheels are large and black, complementing the overall design of the car. The car is parked in a driveway, suggesting that it is in a residential area. The car is the main focus of the video, and there are no other objects or people in the scene. The video is a simple yet elegant showcase of the car, highlighting its design and features."
+    ],
+    "video_ids": [
+        "Xx1kFtMVs8U_20_179to317"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black Porsche sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video consistently shows a black Porsche sports car parked on a driveway, matching the core description. The car's distinctive design, including its headlights, wheels, and body shape, are clearly visible and consistent with a Porsche sports car. There are no elements that contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts shows a black sports sports car, in a driveway. which the description description of The car's sleek features, including its sleek, wheels, and body shape, align clearly visible and align with the Porsche sports car. There are no additional in contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..476f82f829fed0c96d0036944d9e3fed5f27a79b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:644b539e8fa0aee55e202d7eb5233838a64b40a12f0e79f392cf25261c55e41b
+size 225013
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..b66de165601468082b6f1dbbbcfe90163121c64e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman named Sonia Jones, who is the main subject. She is seen in a grey shirt, sitting in a room with a brick wall and a statue in the background. The room has a warm and inviting atmosphere, with a window that lets in natural light. Sonia appears to be engaged in a conversation or an interview, as she is looking to her left with a thoughtful expression on her face. The video seems to be a casual, yet professional, discussion or interview, possibly related to her work or personal life."
+    ],
+    "video_ids": [
+        "FWZBuc13Zuw_4_17to189"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman named Sonia Jones in a grey shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman with blonde hair wearing a grey shirt, and a lower-third graphic identifies her as 'SONIA JONES'. The description matches the core elements of the video without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a woman wearing long hair wearing a grey shirt, which she ring body graphic identifies her as SoniaSIA JONES'. The setting matches the core elements of the video, any.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a6451f76fcc08cd43f70dd34f7f6025c571848fe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:306e97838379836b29cc3336fdbffa48c4166a93108df9c1e562310682e2c7b1
+size 86827
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..903cfa908360136838bd16fea84da450440ea5d8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up portrait of a woman with a focus on her face and upper neck. She has a fair complexion with freckles, and her skin appears to be well-moisturized and glowing. Her makeup is natural, with a subtle contour and a light pink lip color. Her eyebrows are well-groomed and filled in, and her eyelashes are long and voluminous. She has a small gold earring in her left ear. Her hair is styled in a messy bun, and she is looking directly at the camera with a neutral expression. The lighting in the video is soft and even, highlighting the woman's features without creating harsh shadows. The style of the video is realistic and naturalistic, with a focus on the woman's beauty and the details of her appearance."
+    ],
+    "video_ids": [
+        "9PQHWIN4xFw_19_0to128"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Woman with fair complexion, freckles, moisturized skin, natural makeup, small gold earring, messy bun hairstyle.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with fair complexion and visible freckles, her skin appears moisturized, and she has natural makeup with defined eyebrows and lashes. She is wearing small gold earrings and her hair is styled in a messy bun. All elements of the description are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a woman with fair complexion, fre freckles, which skin appears moisturized, and she is natural makeup. a eyebrows and a. She is wearing a gold earrings and her hair is styled in a messy bun. The these of the description are present represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ef586d8841495629bda3d5b110bf96dba50596ff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1199-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef55c48e4d24d39469903eb2363f194f2c43b00d900fd52d109995487a5f13f1
+size 149515
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..ee221634cbb3e57a4aab83e257de0a8a79e7f620
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse of a car's interior, showcasing the transformation from a standard car to a racing car. The first frame shows the car's interior with a standard steering wheel and seats. The second frame shows the car's interior with a racing steering wheel and racing seats. The third frame shows the car's interior with a racing steering wheel, racing seats, and a roll cage installed. The style of the video is a time-lapse, capturing the transformation of the car's interior over time. The video is informative, providing a visual representation of the process of converting a standard car into a racing car."
+    ],
+    "video_ids": [
+        "4aYGC-bMC7o_33_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Standard steering wheel, standard seats, racing steering wheel, racing seats, roll cage? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a car interior with a racing steering wheel (black with red accents and central logo), racing seats (black with red accents, low-profile design), and a visible roll cage structure (metal bars behind the seats and door frames). These elements align with the description of racing-oriented features, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a car interior with a steering steering wheel andtop with red accents) a logo) racing seats (black with red accents), likely and,), and a roll roll cage (.yellow framework around the seats). steering panels). The elements align with the ' of ' seats features, while the 'r(s)' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0315c36c9993347f82e72571d55913fcbcc01dd5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d47485172e60729ccd50d1671d1298bfb4de3f746cbb4ef58f3edd5dffd125de
+size 222456
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..38e64defd2d27a4aea23e715214a4db8dea7fb70
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a conversation between two men, one with blonde hair and the other with gray hair. They are both dressed in suits and ties, indicating a formal setting. The man with blonde hair is speaking, while the man with gray hair is listening and smiling. The background is blurred, but it appears to be an indoor setting with a window. The style of the video is a close-up shot, focusing on the two men and their interaction. The lighting is bright, highlighting the subjects and creating a clear image. The video captures a moment of dialogue between the two men, suggesting a professional or political context."
+    ],
+    "video_ids": [
+        "FDv6GFL3XqM_23_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men in suits and ties, one with blonde hair, the other with gray hair.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men in suits and ties, one with blonde hair (Donald Trump) and the other with gray hair (Vladimir Putin), which matches the description. The background and additional elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men in suits and ties, one with blonde hair andon Trump) and the other with gray hair (Mikeladimir Putin). which matches the description provided The setting appears setting elements do not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..19578c4d3a2726a98042a3edbebada91b5caa2c1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:168b6700f879d96b49e6fbc580021284988ec5e21d50cbcccc4681a5ba2c4d94
+size 235898
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..03f0cb608e821eb37c88fd959aa242273ca8c2b9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are seen discussing a new Jurassic World movie. They are holding toy dinosaurs, with the man holding a white and orange dinosaur and the woman holding a yellow and orange dinosaur. The background features a green screen with a forest scene. The video is a news segment from NBC News, with the title \"The Way of the Dinosaurs\" displayed at the bottom. The style of the video is informative and casual, with the two individuals engaging in a conversation about the movie."
+    ],
+    "video_ids": [
+        "OEgK9I5HMEE_12_0to130"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: man, woman, toy dinosaurs (white and orange, yellow and orange), green screen (forest scene), title 'The Way of the Dinosaurs'? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a woman holding toy figures that are white and orange, and yellow and orange, respectively. The background is a green screen depicting a forest scene. The title 'The Way of the Dinosaurs' is displayed on the screen, along with the names of the actors and the show's branding. All core elements from the description are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man and a woman holding toy dinosaurs that resemble clearly and orange and and yellow and orange, respectively. They background is a green screen depicting a forest scene. The title 'The Way of the Dinosaurs' is present at the screen, which with the text ' the individuals, the video's title. The elements elements of the description are present in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ecccb5c19847b6e2a3619bf70aaf0867cc11000e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e6f59b961be9b2293557aacbfd417338f8ec7d75997c08f4d06b3386a74db205
+size 206598
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..011b3cee9d990cd1454e308df99ccc922ad18031
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a slice of pizza being taken from a larger pizza. The pizza is topped with green vegetables, likely spinach, and has a golden-brown crust. The slice is being lifted from a white paper plate, which is placed on a wooden table. The person taking the slice is not visible in the video, but their hand is visible holding the slice. The style of the video is simple and straightforward, focusing on the action of taking a slice of pizza. The video does not contain any additional elements or background details, keeping the focus solely on the pizza and the action of taking a slice."
+    ],
+    "video_ids": [
+        "AnFeMq_mWvc_0_0to101"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A slice of pizza, green vegetables (likely spinach), golden-brown crust, white paper plate, a hand holding the slice? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a slice of pizza with green vegetables (likely spinach), a golden-brown crust, served on a white paper plate, and a hand holding the slice. The description matches the visual content without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a slice of pizza being green vegetables,likely spinach), a golden-brown crust, and on a white paper plate. and a hand holding the slice. The presence matches the core elements of any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8bd5b02e1a3663ba1050d870a597baa05b699ceb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dce60e4dca0dc7ecea12c7e4e98c22608153abdc6babedbccfe43c10bb873ba6
+size 159786
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..7ca4c9dcd9e36f4cedfa6b96043bfd5e86ee8bbe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a scene of a group of soldiers in a military setting. The central focus is on two soldiers engaged in a conversation. The soldier on the left is wearing a camouflage uniform and a black cap, while the soldier on the right is wearing a white t-shirt. The soldier in the camouflage uniform is speaking to the soldier in the white t-shirt. The other soldiers in the background are also wearing camouflage uniforms and are standing at attention. The setting appears to be outdoors, possibly in a training or deployment area. The style of the video is realistic, capturing the everyday interactions and activities of soldiers in a military environment."
+    ],
+    "video_ids": [
+        "4jDe-oU_QW4_51_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two soldiers engaged in a conversation, other soldiers standing at attention in the background, a soldier in camouflage uniform, a soldier in a white t-shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two soldiers engaged in conversation, with one in a camouflage uniform and another in a white t-shirt. Other soldiers in camouflage uniforms are visible standing in the background, appearing to be at attention. All elements described in the condition are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts two individuals engaged in a, with one soldier a camouflage uniform and the in a white t-shirt. In soldiers in camouflage uniforms are standing in at the background, maintaining to be at attention. The elements in in the condition are present in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..363daccce1020b2f69ee69ac8823f90bbb346414
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bdc7f411c2cb9bdb48bd47c8e7500622c66b7d2ca700204cf5462e57cc18db29
+size 187774
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..119f2062f9a5d5795dfd327032a6c3fbf0fb5f97
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman sitting at a table in a library. She is wearing a pink jacket and a white shirt. Her hair is dark and curly. She is looking at the camera with a slight smile on her face. The library has a wooden floor and a wooden bookshelf in the background. There are blue chairs around the table. The woman appears to be engaged in a conversation or an interview. The lighting in the room is soft and natural, suggesting it might be daytime. The overall style of the video is casual and relaxed."
+    ],
+    "video_ids": [
+        "2ej7nnuLpA8_9_0to184"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a table, blue chairs, a pink jacket, a white shirt, dark curly hair? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman seated at a table with blue chairs around her. She is wearing a pink jacket over a white shirt and has dark curly hair. All specified elements are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a woman wearing at a table with blue chairs. it. She is wearing a pink jacket over a white shirt, has dark curly hair. The the elements are present and match depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..65882a2b8799ac176e216ec33dd5f5d89571a423
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e6860bdd5e8f70aa2d8e93c5482c0a304bfd124bd862a92c235dd4041bda829
+size 115561
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..9525959ecd935083749628643d55473322f62946
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a bartender is seen crafting a cocktail in a bar setting. The bartender is holding a cocktail glass filled with a yellowish drink, possibly a margarita, and is garnishing it with a sprig of mint. The bartender is surrounded by various bottles of alcohol, including tequila, which is likely the base of the cocktail. The bartender is using a cocktail shaker to mix the drink, and a spoon to stir it. The bar counter is well-stocked with a variety of bottles, indicating a wide selection of drinks available. The bartender's focus and attention to detail suggest a professional and skilled approach to crafting cocktails. The overall style of the video is a close-up shot of the bartender and the cocktail-making process, emphasizing the bartender's actions and the drink being made."
+    ],
+    "video_ids": [
+        "hO5-H9X-cH8_0_0to162"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bartender, cocktail glass, sprig of mint, cocktail shaker, spoon, bottles of alcohol (including tequila)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a bartender (a woman in a black top), a cocktail glass with a drink and a sprig of mint, and multiple bottles of alcohol including tequila (labeled 'Tipsy Bartender Tequila'). Although a cocktail shaker and spoon are not visible in the frames, the core elements requested are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as showing showing a bartender preparingpart person in a black shirt) a cocktail glass with a spr and a sprig of mint, a a bottles of alcohol, whatquila.oneabeled astey Bartender').quila'). The the spoon shaker is a are not explicitly, the frames provided the bartender elements of are present, the depicted.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..79952e24c6d5d2665fdcd09cf1bcd619050d8fbe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5cf4d855f0865dc04e6b19cc1acd84efe5262de5d168ddf325e47f9908f2f06e
+size 229174
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..5be289760dec713621d40a521a61c22ceb36157c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and a surprised expression, sitting in front of a desk with a laptop and a mug. The room has a modern and colorful design, with shelves filled with various items, including a trophy and a figurine. The lighting is bright and vibrant, with blue and purple accents. The man is wearing a black t-shirt and appears to be in a casual setting, possibly a home office or a studio. The overall style of the video is lively and energetic, with a focus on the man's reaction and the colorful background."
+    ],
+    "video_ids": [
+        "Lq3jhd2PQOY_0_59to202"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man with a beard and surprised expression, desk, laptop, mug, shelves, trophy, figurine, black t-shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man with a beard and an expressive face, sitting at a desk with a laptop and a mug. Behind him are shelves with various items, including a trophy and figurines. He is wearing a black t-shirt, and all these elements are clearly visible and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man with a beard who a expression face, sitting at a desk with a laptop. a mug. The him, shelves with various items, including a trophy and aines. The is wearing a black t-shirt. which the these elements match consistent visible and match the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8df1b92ea4cf4844e0b92aeef9a0606603d6e80f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-12-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23e355f00e2c74980c37835ca421300799d3f116ced84f3c93af97ab81ddb679
+size 156652
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..b9bafe8946e72332014fb585741aaac2ea17c620
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a first-person perspective of a vehicle driving down a highway. The dashboard of the vehicle is visible, showing the speedometer and other gauges. The road ahead is clear, with a long stretch of highway leading into the distance. The sky is blue with a few clouds, and the surrounding landscape is lush and green. The vehicle appears to be a modern car, and the dashboard is well-lit, indicating that the vehicle is in operation. The overall style of the video is realistic, with a focus on the driving experience and the surrounding environment."
+    ],
+    "video_ids": [
+        "7VFa-hUo9YM_5_0to117"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Vehicle, dashboard, speedometer, gauges, road, sky, clouds, landscape? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a vehicle's interior with a visible dashboard, speedometer, and gauges. The view through the windshield shows a road, sky with clouds, and surrounding landscape, all of which are explicitly mentioned in the condition. No contradictory elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a vehicle's dashboard, a dashboard dashboard, speedometer, and gauges. The road outside the windshield shows a road, sky, clouds, and a landscape, which of which are consistent mentioned in the condition. The additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f0f9841ae9be0c12e0154b7165e329424c81cf00
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d72b7b7ee467d958a026b1086edce7683042982373a23488d64715b85073d3cf
+size 263064
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..361068cb3ebdac4102a0f4c5bad6d51f51b1ed79
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene desert landscape, with a majestic mountain range in the background. The mountains, with their rugged peaks, are bathed in the warm glow of the setting sun, casting long shadows across the sandy terrain. In the foreground, a small building with a tarp roof stands, adding a touch of human presence to the otherwise untouched wilderness. The sky above is a clear blue, with a few clouds scattered across it, adding depth to the vast expanse. The overall style of the video is tranquil and naturalistic, capturing the beauty of the desert and its surroundings in a realistic and unobtrusive manner."
+    ],
+    "video_ids": [
+        "UWvlkfwpF5c_76_0to192"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small building with a tarp roof in the foreground.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a small building with a tarp roof in the foreground, partially obscured by trees and bushes. The structure is visible throughout the frames, and while there are additional elements like trees and distant mountains, they do not contradict the core description of the building with a tarp roof.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a small building with a tarp roof in the foreground, which obscured by the. bushes. The structure is consistent and the frames, fulfilling it there are additional elements like mountains and bushes mountains, they do not contradict the core description of the small with a tarp roof.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..eb3448f9b5a4b612eb8092e7e0a4749321b7bdc7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d490592efd2e83540b553a636ece528df495ce1c26b9f1a4d91fb8e1f6db363
+size 72367
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..3d85190e101f3763d43902e45f2bdb621b41a77e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a lush tropical forest, teeming with life and color. The camera follows a winding path through the dense foliage, revealing a variety of exotic plants and flowers. The vibrant green leaves of the trees and bushes contrast beautifully with the bright orange and yellow flowers that dot the landscape. The forest floor is a rich brown, covered in fallen leaves and twigs, adding to the sense of a thriving ecosystem. The video is shot in a realistic style, capturing the natural beauty of the forest with stunning clarity and detail. The overall effect is a captivating glimpse into the wonders of nature."
+    ],
+    "video_ids": [
+        "2TdhmjbQS2c_13_0to148"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Exotic plants and flowers, fallen leaves and twigs.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a lush, tropical environment with a variety of exotic plants and flowers, such as orange blooms and large green foliage. The ground is covered with fallen leaves and twigs, contributing to the natural, jungle-like setting. These elements align with the described condition, even though there are also other decorative elements like ropes and carved structures, which do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a lush garden vibrant setting with vibrant variety of exotic plants and flowers, which as the bloss and green green leaves. The ground is covered with fallen leaves and twigs, creating to the natural, serene-like setting. The elements align well the description ', making though the are no some elements elements like a and a figures that which do not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fe5dad7293027981eea69014b7e5e289322ea92c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4bb0227e7c515420297f859dc5316972633f9c088fa96d5e0621f89eccc428b7
+size 350232
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..e1d24b07a0538d7b466b98cc6131ba32015d17c1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a young man sitting inside a vehicle, likely a bus or van, with a window behind him. He is wearing a brown jacket and a baseball cap. The man appears to be in motion, as suggested by his hair being blown by the wind. The window behind him shows a glimpse of the outside world, with a gas station and a car visible. The style of the video is casual and candid, capturing a moment in the man's journey. The focus is on the man, with the background serving as context for the setting."
+    ],
+    "video_ids": [
+        "PY8M7LSwlQk_19_49to247"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man, a brown jacket, a baseball cap, a gas station, a car.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a young man wearing a brown jacket and a baseball cap, seated inside a car. Through the window, a gas station is visible in the background. All core elements from the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a young man wearing a brown jacket and a baseball cap. sitting inside a vehicle. The the window, a gas station and visible, the background, There the elements of the description are present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b55c4441b5a8b51e5805dc3e2c8d236aaf031286
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da63875b5c0ae1356c4e8afef4e22fb6cee378d488925ebea771fa6caca916b2
+size 196964
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..c1f2e9194670aefd1f7c79d6c5fcdd65d5fbd679
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a drone shot of a castle located on a cliff overlooking the ocean. The castle is large and imposing, with multiple towers and walls. The surrounding landscape is lush and green, with a few trees scattered around the castle. The ocean is visible in the background, with waves crashing against the cliff. The sky is overcast, with no visible clouds. The castle appears to be in good condition, with no visible damage or decay. The drone shot provides a comprehensive view of the castle and its surroundings, showcasing the beauty and grandeur of the structure."
+    ],
+    "video_ids": [
+        "W81nMnUapsM_6_0to119"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Castle, multiple towers, walls, trees, ocean, waves, cliff? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a castle with multiple towers, surrounding walls, trees, an ocean, waves, and a cliff. The aerial view captures all these elements as described, with no contradictions in the visual representation.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a castle with multiple towers, walls walls, trees, an ocean, waves, and a cliff. The aerial view provides the these elements, described, with the contradictions or the presence content.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e9da1471dc49dbfd842f22278dae3eccd25cc9cd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba9d565ea48f9d8c660c20456b1a07c4096cf0262e9d62a5875c59bfb97eb221
+size 184301
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..0959858e5dd45c57351b54f0c773f92736f68799
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a pile of marijuana buds. The buds are dense and covered in trichomes, giving them a frosty appearance. The color of the buds ranges from green to purple, with some orange and red hues. The buds are stacked on top of each other, creating a small mound. The background is a plain white surface, which contrasts with the vibrant colors of the marijuana. The style of the video is realistic, with a focus on the details of the marijuana buds. The lighting is bright, highlighting the trichomes and the colors of the buds. The video does not contain any text or additional elements."
+    ],
+    "video_ids": [
+        "TKPFnjJzy44_26_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A pile of marijuana buds, dense and covered in trichomes, ranging in color from green to purple with orange and red hues.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a pile of marijuana buds that are dense and covered in trichomes, as described. The coloration includes green and purple hues, with visible orange and red accents, matching the description. There are no conflicting elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts a dense of what buds that are dense and covered in trichomes, which described. The buds range of shades, purple hues with with some orange and red tones, which the description provided The are no additional elements in would the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c26265b9273c209770d02f707efb2c92f0e1742b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c2f7c92aff4f34849666f1186e068c9099fec1a113dd77869180694e8fc80cd4
+size 124574
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..1442bb4c83e7813fe01eb85094e9a4a50d9b8a4a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man being interviewed by multiple reporters. The man is wearing a gray suit and a blue shirt. He is standing in front of a building with a glass facade. The reporters are holding microphones and recording devices. The man appears to be speaking to the reporters. The building in the background has a modern design. The sky is clear and blue. The man is the main subject of the video. The reporters are the secondary subjects. The building is the background. The sky is the background. The man is speaking. The reporters are recording. The building is made of glass. The sky is blue. The man is wearing a suit. The reporters are holding microphones. The building is modern. The sky is clear. The man is standing. The reporters are standing. The building is tall. The sky is bright. The man is talking. The reporters are listening. The building is glass. The sky is blue. The man is wearing a shirt. The reporters are wearing clothes. The building is tall. The sky is clear. The man is standing. The reporters are standing. The building is made of glass. The sky is blue. The man is wearing a suit. The reporters are wearing clothes. The building is tall. The sky is clear. The man is talking. The reporters are listening. The building is made of glass. The sky is blue. The man is standing. The reporters are standing. The building is tall. The sky is clear."
+    ],
+    "video_ids": [
+        "CZ7faM4aOyM_10_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ['A man', 'Multiple reporters', 'A building with a glass facade', 'Sky']? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man being interviewed by multiple reporters, with a building featuring a glass facade in the background. Although the sky is not prominently visible, the outdoor setting implies its presence, and the core elements described are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man being interviewed by multiple reporters, which a building featuring a glass facade and the background. The the sky is not explicitly visible, it presence setting and the presence. fulfilling the blue elements of in present represented.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4d5de2b54ab47b398420bffac0d65a9848ed318e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6417857d81f6ec0bb4d30d5b1cff028db6e442039dfa66ed18f2ecf85a065f02
+size 233263
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..963adabefbbe30c56dcab556d96ea8286031e33e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a cozy, snow-covered cabin nestled in a forest during winter. The cabin, with its wooden exterior and snow-covered roof, is surrounded by tall, snow-laden trees that create a serene and peaceful atmosphere. The snow blankets the ground, and a wooden fence can be seen in front of the cabin, adding to the rustic charm of the scene. The video is likely a 3D animation or a digital rendering, given the level of detail and the realistic depiction of the snow and the cabin. The overall style of the video is warm and inviting, evoking a sense of comfort and tranquility."
+    ],
+    "video_ids": [
+        "FU47JEZ4TRA_48_0to139"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Cabin, snow-laden trees, snow-covered ground, wooden fence? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a cabin, snow-laden trees, snow-covered ground, and a wooden fence. These elements are prominently featured throughout the video, matching the description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a cabin, snow-laden trees, snow-covered ground, and a wooden fence. The elements are prominently featured in the frames, creating the description provided any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a1074997e7840e3916d78e6e929da83839437298
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-120-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b58ea41eec20794c7e14cff6b01ee232fdf1f9054a2b741716cf6ff9ef1483c
+size 170581
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..e4a4f1cec7eb06d0e0f558d5cd78a9e48f549b46
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is an aerial view of a tropical village nestled in a lush green valley. The village is composed of several thatched-roof houses, each with a unique design and color. The houses are scattered throughout the valley, with some located closer to the camera and others further away. The village is surrounded by dense vegetation, including palm trees and other tropical plants. The video captures the tranquility and beauty of the village, showcasing its unique architecture and the natural beauty of its surroundings. The style of the video is a drone shot, providing a bird's eye view of the village and its surroundings."
+    ],
+    "video_ids": [
+        "cnx41LOX9yk_45_0to147"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Several thatched-roof houses with unique designs and colors.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple houses with thatched roofs, each with unique designs and colors (e.g., white, yellow). The structures are nestled among lush greenery, and their architectural styles vary, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts several houses with thatched roofs, which having unique designs and colors.e.g., some with blue, The houses are surrounded among lush greenery, which the varying styles vary, fulfilling the 'unique(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..54c43ea582f57dd2c092ba75c6b808047f58e46a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1bfc5ecfd37e06d0f64b1c072ca2328cdad41ca6ce1225c98130f68f66643c60
+size 286754
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..f19b1a8accee5935cf045e2fbdb888523d2e5389
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a gardening scene with a focus on two potted plants and a watering can. The watering can is green and has a white ribbon tied around it. The potted plants are placed on a wooden pallet, and the pots are white and blue. The plants are lush and green, indicating that they are well cared for. The watering can is positioned between the two pots, suggesting that it is being used to water the plants. The overall style of the video is simple and straightforward, focusing on the gardening activity without any additional embellishments. The video likely captures the process of watering the plants, highlighting the care and attention given to the plants."
+    ],
+    "video_ids": [
+        "0T0XntBbfZw_0_15to230"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two potted plants and a green watering can with a white ribbon.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two potted plants (one in a white container, one in a blue container) and a green watering can with a white ribbon tied around it. These elements are central to the scene and are accurately represented, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two potted plants,one green a white pot and the in a blue container) and a green watering can with a white ribbon. around its. The elements match present to the description and match consistent represented in fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3c159c89ffaef9c3b5bb74c14ae2e9d2e2c5a2c2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:38dc45441526a3f345f5b08a5752fe344215f17daabb6b6047b142409e98b665
+size 150532
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..d0357ee1ce95db227c134732b58c1e2ae5be6715
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a playful scene in a room with a blue wall and a white radiator. Two suitcases, designed to look like Volkswagen buses, are the main focus. The first suitcase is blue and white, while the second one is green and white. They are positioned on a gray carpet, with the first one slightly ahead of the second one. The suitcases have wheels and handles, suggesting they are meant for travel. In the background, there is a white dresser and a green and white patterned pillow, adding to the room's decor. The overall style of the video is light-hearted and fun, with a focus on the whimsical design of the suitcases."
+    ],
+    "video_ids": [
+        "Ychs5_2awtM_11_0to162"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two suitcases (blue and white, green and white), a gray carpet, a white dresser, a green and white patterned pillow.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two suitcases designed to look like vintage vans, one predominantly blue and white, the other green and white. They are positioned on a gray carpet, with a white dresser visible on the right side of the frame and a green and white patterned pillow behind them. All elements described in the condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows two suitcases, to look like vans vans, one blue blue with white, and other green and white. The are placed on a gray carpet. and a white radiator partially in the right side of the frame. a green and white patterned pillow on the. The elements in in the condition are present in match represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bb33276ada35fc81ff6409cc21b176e58fd0dd57
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d679dadced2705f0cd19d07724b9786e6c1120b0c2dab05b41338d6d7e1834bf
+size 182914
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..e550e0b6feca7eb1b23b795c3a7cbfb1b9466d3a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a promotional advertisement for hammered copper barware. It features a close-up of a copper pitcher and a copper cup, both with a textured, hammered finish. The copperware is placed on a wooden cutting board, which also holds fresh limes and mint leaves, suggesting a refreshing beverage. The background is blurred, but it appears to be a kitchen or bar setting. The style of the video is clean and modern, with a focus on the copperware's texture and the fresh ingredients, likely to evoke a sense of luxury and sophistication. The video is likely to be used in a catalog or online store to showcase the product's design and potential use."
+    ],
+    "video_ids": [
+        "5UXIzzeGKaY_1_0to185"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Copper pitcher, copper cup, wooden cutting board, fresh limes, mint leaves? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully features a copper pitcher (shaker), copper cups (mugs), a wooden cutting board with mint leaves, and fresh limes. These core objects are clearly visible and match the description. Additional elements like a jigger and a tray are present but do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a copper pitcher,likelyiny) a cup,mugs), a wooden cutting board, fresh leaves, and fresh limes. The elements elements are clearly visible and match the description provided The elements like the knifeigger and a knife are not but do not contradict the core and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..28f15a736a6bae5437c8aac245cec1ff3f9ac1f7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9b9f22a7e58e0fb54308918bd75dc3533307cbf85518fceb3a0304163ab82853
+size 102863
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..edfe7fd319479e4cc1ed73fe7d63ff1fe844c6d0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up of a tiger's face in a dense forest. The tiger, with its distinctive black stripes, is the main focus of the video. The camera is positioned in such a way that it captures the tiger's face in detail, from its eyes to its nose. The background is filled with lush green foliage, creating a natural and serene setting. The video is shot in a realistic style, with the camera remaining still, allowing the viewer to fully appreciate the beauty and majesty of the tiger in its natural habitat. The overall composition of the video emphasizes the tiger's presence in the forest, making it a captivating and engaging visual experience."
+    ],
+    "video_ids": [
+        "U1LhgusHlA8_79_49to226"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A tiger with distinctive black stripes.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a tiger with distinctive black stripes on its orange fur, which matches the description. The tiger's facial features, including its eyes, nose, whiskers, and striped pattern, are visible and consistent with the expected appearance of a tiger. There are no elements in the video that contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a tiger with distinctive black stripes, its face fur, which is the description of The tiger's facial features, including its eyes, nose, anders, and the pattern, are prominently and consistent with the typical appearance of a tiger. The are no additional in the video that contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..552313713606bea98888bf0f5815ee23d0b334cb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8d6012e6222bae8d1380a85961376eaf5ba2fbacfeb02c6f33543bee639df1e7
+size 295056
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..eecfed5fb7382f9d841b04c204ff39cd1bf85137
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a blue jersey, presumably a soccer player, with a focused expression on his face. He is looking to the side, possibly at an opponent or teammate. The background is blurred, but it appears to be a soccer stadium filled with spectators. The style of the video is a sports action shot, capturing the intensity and concentration of the player during a game. The focus is on the player, with the background serving to emphasize the competitive atmosphere of the event."
+    ],
+    "video_ids": [
+        "MbzRUFkqeqQ_36_67to203"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a blue jersey (presumably a soccer player)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue jersey, which is consistent with the description of a soccer player. The presence of the FC Bayern M\u00fcnchen logo and blurred stadium background supports the context of a soccer setting without contradicting the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue jersey, which is consistent with the description of a soccer player. The jersey of a jersey logo logo logo on the spectators background further the context of a soccer match. contradicting the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cedf0949e8b6dd9a90d75e26e82ec5b23403e146
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:93ebdc66965dbb2b175529162ddad3c1b019d6c47fe96a18a1e0f8d0f4ed4978
+size 155800
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..5077dc1b9d508dc103a9351ceb29fd2eb09e6b7c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a scene of a motorcycle procession on a suburban street. The procession is led by a police officer on a motorcycle, followed by several other motorcycles. The motorcycles are moving in a line, one behind the other, and they are all moving in the same direction. The street is lined with houses and parked cars, and there are trees and power lines visible in the background. The sky is clear and blue, suggesting that the weather is sunny and pleasant. The video is shot from an aerial perspective, providing a bird's eye view of the procession and the surrounding area. The style of the video is realistic and documentary, capturing the event as it unfolds without any artistic embellishments."
+    ],
+    "video_ids": [
+        "bZTCXfss5qs_22_0to131"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A police officer on a motorcycle leading a procession of several other motorcycles.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a police officer on a motorcycle leading a procession of several other motorcycles, which matches the described condition. Additional elements like cars and houses are present but do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a scene officer on a motorcycle leading a procession of several other motorcycles, which align the description '. The elements such the parked houses in present but do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1188f8d501a09406069b0e007e330303806c2fa9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:51ef7ec8c1e80d693f18e7ec2b67b95650b38cffb9a194d582c846fd18038ada
+size 197120
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..4af8001551ba97dc9d17cc6a95920832ecfec1c8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a puffin swimming in calm, rippling water. The bird's distinctive black and white plumage is clearly visible, along with its bright orange beak and feet. The puffin is seen paddling vigorously with its wings, creating small splashes as it moves through the water. The background consists of gentle waves, suggesting a serene coastal environment. The BBC logo is visible in the top left corner, indicating the source of the footage. The camera remains steady throughout the sequence, focusing on the puffin as it swims from left to right across the frame. The overall scene conveys a sense of tranquility and natural beauty."
+    ],
+    "video_ids": [
+        "196033bd3cb251b74cf973ae87e5176877fe31835a0502c48a8102923225a975"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A puffin, gentle waves in the background, BBC logo in the top left corner.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a puffin swimming in gentle waves, which matches the description. The BBC logo is visible in the top left corner throughout the video. There are no elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a puffin swimming in water waves, which ful the ' of The BBC logo is visible in the top left corner, the video, The are no additional in contradict the given description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0a71b69942d63ccdfb43598d1f41164d3651f59e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1200-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54731761183f1f2893354880f2ffc0c690c682527fa134b6ad4556c5c9dacec7
+size 265048
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..4c7f289dd9930974c3fe49562e1b0c28ce5aca0b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a camouflage outfit, including a jacket and a hat, standing next to a white truck. He is wearing a backpack and has a gun slung over his shoulder. The truck is parked on a dirt road, and the man appears to be looking off into the distance. The time displayed on the screen is 1:55 PM. The style of the video is realistic and seems to be a documentary or a real-life scenario. The focus is on the man and his actions, with the truck and the surrounding environment serving as the backdrop."
+    ],
+    "video_ids": [
+        "3pfKs_LCPvw_18_0to140"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a camouflage outfit (jacket and hat), a backpack, and a gun slung over his shoulder.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed in a camouflage outfit, including a jacket and hat, carrying a large camouflage backpack. He is also holding a bow (which can be interpreted as a 'gun' in a hunting context), slung over his shoulder. The core description is accurately fulfilled, with no significant contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing in a camouflage outfit, including a jacket and hat, which a backpack backpack-pattern. He is also wearing a gun andwhich can be considered as a gungun' in the non context) slung over his shoulder. The setting elements of largely represented by with the significant contradictions or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fff57d98109e09a6927513fbcbfa65d52e1801f7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4fad1f0c1264a202f575425e5214ea0627b6c646867145ea88e21ed7dbddd84c
+size 211466
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..a45371675c3ad0dc83cf4ee2b8fabb161d5d5a79
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a red BMW SUV on display at a car show. The car is positioned on a reflective floor, with a cityscape in the background. The vehicle is the main focus of the video, with its sleek design and shiny exterior reflecting the lights of the showroom. The cityscape in the background adds depth to the scene, creating a contrast between the urban environment and the luxury vehicle on display. The video captures the essence of the car show, highlighting the allure of the BMW SUV and the excitement of the event."
+    ],
+    "video_ids": [
+        "5Ue8CW3uL74_13_0to181"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red BMW SUV? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red BMW SUV, specifically the BMW X2 model, as indicated by the badge on the front. The vehicle is prominently displayed, and its features, including the iconic BMW grille, headlights, and overall SUV design, are consistent with the description. The background elements, such as people and cityscape, do not contradict the core object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a red BMW SUV, which a BMW X5 model, which indicated by the visible on the front. The vehicle is positioned displayed in and the design, such the grille BMW grille and headlights, and overall design design, are visible with the description. The background,, such as the and thescape, do not contradict the core description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f2bc47d990bfd88012b3c9330458d2e6b82ec737
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9876de5a720fa0b867a63bb4d3cb127ba1fedef3a67b911be42af3e2197c0c12
+size 137206
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..11a57af3c621bbbddf4e717a067593f9f54deaef
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are seen laughing together on a television show set. The man is wearing a white shirt and glasses, while the woman is dressed in a colorful floral top. They are standing in front of a backdrop that features a cartoon character and the words \"Plaza Ambush Makeover\". The overall style of the video suggests a light-hearted and fun atmosphere, with the two individuals enjoying a moment of shared amusement."
+    ],
+    "video_ids": [
+        "PhTLP9zh8aM_2_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman sitting next to each other, engaging in conversation and smiling. Their presence and interaction fulfill the 'Object(s): A man and a woman' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man and a woman standing closely to each other, which in a. smiling. The attire and interaction fulfill the 'Object(s)' A man and a woman.' condition. any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b3fe0184ad4bdd3a5f2a116dbf393e694b290364
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59bd993d440fad692c80beaeac4c9cd3e06a9d6a3fcde06d2e8d5f2261b0b4b6
+size 304545
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..c404ef36ec44b7ee540fbe9a321d4461de6c812f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video begins with a close-up view of a cylindrical object that appears to be a piece of wood or bark, heavily covered in a greenish-yellow substance. This substance is likely mold or some form of organic growth, as it has a slimy texture and is dotted with small yellow specks. The surface of the object is rough and uneven, with some areas appearing charred or burnt, suggesting it might have been exposed to heat or fire at some point. As the video progresses, a small black insect, possibly a beetle or a similar creature, emerges from the center of the object. It moves slowly across the surface, exploring its surroundings. The insect's movements are deliberate and cautious, indicating it may be searching for food or a suitable environment to lay eggs. The camera remains steady throughout, focusing on the insect and the object, providing a detailed view of the interaction between the two. The background is a plain, dark gray, which helps to highlight the colors and textures of the object and"
+    ],
+    "video_ids": [
+        "798f043187cfb2b1880032ac554c799372d384f54989b31d8954d590d1068b6c"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Cylindrical object, small black insect? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a cylindrical object with a textured, multi-colored surface, and a small black insect is visible on its side. The core elements described in the condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a cylindrical object covered a textured surface colorful-colored surface that which a small black insect is visible on the surface. The cylindrical elements of in the condition are present, match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a0e3ee73437a7b652e9c2387ee259f82e20bbb3c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1208065e6743f5e5a56a51ef21ee1d27019d94e6fa4b3abfad24414f511bce79
+size 265155
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..c54f7a6dca7c23a616d62893722c1d5a2755c1b2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a high-speed boat journey across a vast body of water. The boat, painted in striking shades of orange and white, cuts through the water, leaving a trail of white foam in its wake. The boat's design is sleek and modern, with a curved hull that slices through the water with ease. The water around the boat is a deep blue, reflecting the clear sky above. The boat is moving at a high speed, as evidenced by the white wake it leaves behind. The video is taken from a high angle, providing a bird's eye view of the boat's journey. The overall style of the video is dynamic and action-packed, capturing the thrill of high-speed boating."
+    ],
+    "video_ids": [
+        "aebVBldEHm0_36_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A high-speed boat painted in striking shades of orange and white, with a sleek and modern design.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The boat in the video is clearly painted in striking shades of orange and white, with a sleek and modern design. The high-speed nature is implied by the wake and motion. No elements contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video in the video is indeed high in striking shades of orange and white, which a sleek and modern design. The high-speed movement of evident by the wake and the of The additional contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ed8f8857357ee3e1ecfa2bcb933f31f4aadd923d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d9e85c34bd34598edcafb1cd5df86eeabf588800cdc7beaf06ec506659c6cc4
+size 380515
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..3d7fd66f874856691605199ee94a8f63d592500d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a series of clear plastic bags filled with water and small fish, likely betta fish, arranged neatly on a blue textured surface. The bags are sealed at the top with a red twist tie, and some condensation is visible on the outside, indicating a cool environment or recent handling. The fish inside the bags are mostly red with some having darker fins, and they are swimming slowly, occasionally moving their tails. The background includes a white towel draped over the edge of the surface, suggesting a setup for transporting or displaying the fish. The lighting is natural, possibly from sunlight, which highlights the clarity of the water and the vibrant colors of the fish. There are no significant changes or movements throughout the video; it focuses on the static arrangement of the bags and the gentle activity of the fish within them."
+    ],
+    "video_ids": [
+        "cf4d9793188f734803dc6ec4f5f206324a59f6be21c4843d00216d0ef03a5c3d"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Clear plastic bags, small fish (likely betta fish), red twist ties, condensation on the bags, white towel? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple clear plastic bags filled with water, each containing a small fish (consistent with betta fish), and condensation is visible on the bags. A white towel is also visible on the right side of the frame. While red twist ties are not clearly visible, the core elements described are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows clear clear plastic bags containing with small and containing containing small small fish,likely with theta fish due secured securedensation is visible on the bags. The red towel is also present beneath the surface side of the frame. The the twist ties are not explicitly visible in the presence elements of in present, match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e37a6c965c5f3cb762aa008f850e742b0f4fc335
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f7a053a4734fb272387a387e1a170941186d4cdd3d269a36a7c20131a12e132b
+size 175252
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..d50de648e5134c9f01ffee50b0b615dc6903e573
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a stunning aerial view of a coastal town nestled between towering mountains and a deep blue fjord. The town is densely packed with houses and buildings, their colors muted against the vibrant green of the surrounding landscape. The fjord is a deep blue, reflecting the clear sky above. The mountains rise majestically in the background, their peaks dusted with snow. The video captures the beauty of the landscape and the tranquility of the town, offering a unique perspective on this picturesque location."
+    ],
+    "video_ids": [
+        "7lvXbfNBIQg_167_562to702"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Houses and buildings in the town, mountains, a fjord, and the sky? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing houses and buildings in the town, towering mountains with patches of snow, a fjord with deep blue water, and a bright blue sky with scattered clouds. All elements described are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing houses and buildings in a town, mountains mountains, snow of snow, a fjord with deep blue waters, and a bright blue sky. no clouds. The these are in present and contribute depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..079593b19baf2ed8a8f5e8affa773c6638c79290
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3ad84d1159e565802f677ea9cf89d378694c9be71bf36050c07d643b252a107f
+size 191528
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..cb8fad49102525de2ee07268025efb6e8d0e40ce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man sitting on a stool in a bustling market, engrossed in his work on a calculator. He is dressed in a white shirt and a blue apron, indicating his role as a vendor. The market is filled with various items, including books and boxes, suggesting a diverse range of products available for sale. The man's focused demeanor and the lively atmosphere of the market create a dynamic scene that encapsulates the essence of daily life in a bustling marketplace."
+    ],
+    "video_ids": [
+        "0dAuHLEXLyM_162_258to391"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a stool, a calculator? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man sitting on a stool while holding a calculator. These three objects are prominently featured and accurately match the 'Object(s)' condition described. Additional elements in the background (e.g., other people, shelves, boxes) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man sitting on a stool, using a calculator. The elements elements ( present featured in match represent the descriptionObject(s)' condition described. The elements like the video,such.g., boxes boxes, boxes, and) do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..21f8f2f162cad6b9b8cf6c2fc89b7190d4060ee1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1201-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1250f8ee675122592894c2e054e4c7ada2069dffced61882a5c757cea9b4fa1
+size 139283
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..3cb88d53d0c29de310b17598521b718119c5772a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases the ID.3 electric car by Volkswagen, highlighting its affordability at $33,690. The car is displayed in a vibrant, colorful setting with a purple and blue background, featuring a neon light design. The car itself is a hatchback with a sleek design, featuring a red and blue color scheme. The video emphasizes the car's affordability and design, making it an appealing choice for electric car enthusiasts."
+    ],
+    "video_ids": [
+        "EeVUJ9q2jTA_74_24to152"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ID.3 electric car (hatchback with a sleek design)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video displays a Volkswagen ID.3, which is an electric hatchback with a sleek, modern design. The car's shape, proportions, and styling elements (such as the streamlined body and aerodynamic lines) align with the description of a sleek hatchback. The video does not contradict this core description, even though it includes additional elements like text overlays and a stylized background.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a Volkswagen electric.3 electric which is an electric carback with a sleek design modern design. The car is features, including, and the are,such as the front body and theodynamic lines) align with the description of an hatch electricback. The vibrant does not contradict the description description, and though it includes additional elements like a and and a stylized background,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4bef9ac1a5399a48b2b44328908fbf36f7a0204a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c53cb3b0fe231ecd346fca3590b6c2b283cc121e2a941faf2067db02b7a47437
+size 200084
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..d79f4598fad5100f478a0c9b7463b34edd20c067
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with dark hair and glasses, wearing a plaid shirt. He is seated in an office setting with a whiteboard in the background. The man appears to be engaged in a conversation or presentation, as he is looking to his left with a focused expression. The office environment suggests a professional setting, possibly a meeting or a discussion. The style of the video is candid and natural, capturing a moment in the man's workday."
+    ],
+    "video_ids": [
+        "1cs_pL7tJWw_5_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man with dark hair and glasses, wearing a plaid shirt, seated and looking to his left with a focused expression.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with dark hair and glasses, wearing a plaid shirt, seated and looking to his left with a focused expression. The background elements, such as posters and a yellow object, do not contradict the core description and are consistent with an office or creative workspace setting.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with dark hair and glasses, wearing a plaid shirt, seated and looking to his left with a focused expression. The description is, such as the and a window object, do not contradict the core description and are acceptable with the indoor or casual workspace setting.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..13d7b9455ff716c7e3e18da218ae3a75776acacb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:299badab6e4b4647ba3835f18203e017e5410e276f95ece5e83bec198d5d2e2e
+size 147213
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..c0e460379bdc2ea709087f796f1066d9b8e6eec4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a dessert being prepared on a white plate. The dessert consists of a layered structure with a base of a golden-brown biscuit, topped with a creamy white filling, and a final layer of a golden-brown biscuit. The dessert is garnished with a sprig of mint and a few slices of banana. The style of the video is a simple, straightforward food preparation video, focusing on the dessert and its ingredients. The camera angle is slightly above the plate, providing a clear view of the dessert and its garnishes. The lighting is bright, highlighting the colors and textures of the dessert. The video does not contain any text or additional elements, focusing solely on the dessert and its preparation."
+    ],
+    "video_ids": [
+        "YANTpX44ZEs_31_0to116"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A layered dessert, a sprig of mint, a few slices of banana? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a layered dessert consisting of two biscuit-like layers with a creamy filling and banana slices in between. There is also a sprig of mint and additional banana slices arranged around the dessert on the plate, matching the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a layered dessert with of w typesuit-like layers with a creamy filling in banana slices in between. There is a a sprig of mint placed a banana slices on on the top, the plate. which the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..320a3540fd465724e5d8aefab1e11a74df88c85b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:110964fae0548a6aeeb04f19812b3f1497c5dbab1be95075e5fc5e3f726870f3
+size 187502
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..99ffbc4a9cc623239f16672b45800c5acfa769a8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a large sailboat gliding across a vast body of water. The boat, painted in a pristine white, is adorned with three tall masts, each supporting a large sail that billows in the wind. The sails, a mix of white and yellow, add a vibrant contrast to the boat's white hull. The boat is moving towards the right side of the frame, leaving behind a trail of ripples on the water's surface.  The backdrop of the scene is equally captivating. A majestic mountain range stretches across the horizon, its peaks reaching up towards the sky. The mountains, cloaked in a lush green, provide a stark contrast to the clear blue sky above. The sky is dotted with fluffy white clouds, adding a sense of depth and scale to the scene.  The video is shot from a distance, allowing the viewer to take in the entirety of the scene. The perspective is such that the boat appears to be moving towards the viewer, creating a sense of anticipation and adventure. The overall style of the video is one of tranquility and beauty, capturing a moment of calm and serenity in the midst of nature's grandeur."
+    ],
+    "video_ids": [
+        "f1PoCTds2-Y_36_0to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large sailboat, three tall masts, sails, mountains, sky, clouds? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large sailboat with three tall masts, fully rigged sails, mountains in the background, and a sky with clouds, all of which match the described conditions. The elements are clearly visible and do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a large sailboat with three tall masts and sails unf with, and in the background, a a sky with clouds. which of which align the description elements. The presence in present present and do not contradict the given.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..85415215e0026899e181b298f8eb7b35ef7a6fad
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab6ba05a3caa57681a5d146fe2323a933560631240bc5c9fd09574a8d5df0ee9
+size 201865
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..729d22befc77ddbbd06ae8e8261c8087fa7d191f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a sleek, futuristic-looking sports car in a garage. The car is predominantly black with yellow accents and a green stripe running along the bottom. The car's design is aerodynamic, with sharp angles and curves that give it a high-tech appearance. The car is parked in a garage, and the lighting highlights its glossy finish. The garage itself is spacious, with a concrete floor and high ceilings. The car is the main focus of the video, and its design and color scheme are striking. The video is likely a promotional or showcase video for the car, highlighting its design and features."
+    ],
+    "video_ids": [
+        "c6MArvqEjLs_35_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A sleek, futuristic-looking sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a sleek, futuristic-looking sports car with aerodynamic design elements, carbon fiber accents, and a distinctive logo, matching the description. The presence of another car in the background does not contradict the core description of the main subject.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a sleek, futuristic-looking sports car with aodynamic design elements, a fiber accents, and a distinctive color on which the description of The setting of the car in the background does not detr the core description of the main object.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..de79234d0bbc0d5c6f5a4c6386191779d4e8b208
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0dcb3e28f7013c045698d22706ae56c3cb7081f567b2e03d6b6522457b77989
+size 70777
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..7acf26cfefa70223c6c1d74511220a63da0d951d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car from the perspective of the driver's seat. The car has a modern design with a sleek dashboard and a large touch screen display in the center console. The steering wheel is on the right side of the car, indicating that the car is designed for driving on the left side of the road. The car's interior is well-lit, with the dashboard and center console illuminated by the car's lights. The car appears to be stationary, as there are no visible signs of motion or activity outside the car. The style of the video is a straightforward, unembellished documentation of the car's interior, with no additional context or narrative provided."
+    ],
+    "video_ids": [
+        "AdWKZvpGyP8_21_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel (on the right side), dashboard, touch screen display, center console? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a steering wheel positioned on the right side, a dashboard with integrated touch screen displays, and a center console with controls and a gear selector. All core elements described are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a steering wheel on on the right side, which dashboard, a controls screen display, and a center console with a. a gear shift. These these elements of in present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..aec3028b5b8e52584ef4864349223f783ca3b69a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52d3dd9c99742865563b69abc16e669d62669c30dc426d474309d68cdd0d8321
+size 246492
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..8e5a04301abe40e3817a36281c83f1350a4272f5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a garden bed with various plants growing in it. The plants are in different stages of growth, with some being larger and more mature than others. The garden bed is filled with soil, and there are visible signs of watering and care, such as a hose and a watering can. The plants are a mix of leafy greens and other vegetables, and they are arranged in a neat and orderly fashion. The garden bed is located outdoors, and the lighting suggests it is daytime. The style of the video is a straightforward, non-narrative depiction of a garden, focusing on the plants and the care they receive."
+    ],
+    "video_ids": [
+        "p9nUsTG9Uhw_0_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Various plants including leafy greens and other vegetables.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a garden bed with various plants, including leafy greens such as green and red cabbage or lettuce varieties, and smaller radish plants. These plants are arranged in rows and appear healthy and well-tended, fulfilling the 'Various plants including leafy greens and other vegetables' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts various variety scene with various plants, including leafy greens, as lettuce leaf yellow leaf, lettuce,. which other plantsish-like. The plants are growing in a, are to, well-mended, which the conditionObject plants including leafy greens and other vegetables' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b24b1cc08fe7e0181d5cc10201abb7fc62ef3e73
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1242216b3b9ee2a43b3f59fec66e73e126962db2da5a8a2ea858225577bec00f
+size 274579
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..fea1d9aa8782799b8de61014f116b2a51a96fb9a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a brightly lit room with a red and white striped wall. She is wearing a red and white striped sweater and is laughing, her hand on her head. She is standing in front of a potted plant and a shelf with various items on it. The room has a warm and inviting atmosphere. The woman's laughter and the bright colors of her sweater and the room create a cheerful and lively scene."
+    ],
+    "video_ids": [
+        "HDr5PDwUGBE_22_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a potted plant, a shelf with various items on it.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman in a striped shirt, a potted plant in the background, and a shelf with various items (such as cups and a pot) visible on the left side of the frame. These elements align with the described 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a woman wearing a red red, a potted plant in the foreground, and a shelf with various items onbooks as books and books book) on. it right side of the frame. The elements match with the description 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..410c59314f3138ba7a5d8b2010218058c644f59f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1202-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1239616659c367e359eecad7349cac2a5e5faedb4b70c6ae08b1b368785f3a92
+size 238217
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..0fab02f599d82412fab75ed9de14dcfa7ea0bccc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the tranquil beauty of a beach scene. The first frame shows the vast expanse of the ocean, with waves gently lapping against the shore. The second frame reveals the sandy beach, where the waves have left behind a few small ripples. The third frame shows the ocean again, with the waves beginning to recede. The overall style of the video is serene and calming, with a focus on the natural beauty of the beach and the ocean."
+    ],
+    "video_ids": [
+        "AF0wCDjsxQE_163_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Waves, shore, ripples.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows waves rolling onto the shore, creating ripples as they interact with the sandy beach. These elements are central to the visual content and align perfectly with the described conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts waves crashing towards a shore, creating ripples in they interact with the sand beach. The elements are central to the description content and align with with the ' '.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..64c09b4b4baa28dee37a5349e95a900c5b52214d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8bf56945db70b28497777bd01aa6483116e1fbde3b8c942c0a963cd4f8b63ac1
+size 362938
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..a95e2437f2dd4f67a331e537c82f82d876d408b4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car, focusing on the dashboard and the steering wheel. The dashboard is made of wood and has a luxurious appearance. The steering wheel is black and has a leather texture. The car's radio is prominently displayed in the center of the dashboard. The radio has a digital display and a keypad for input. The car's interior is well-lit, with the dashboard and steering wheel being the main focus of the video. The style of the video is a close-up shot, focusing on the details of the car's interior."
+    ],
+    "video_ids": [
+        "cjFEkj8iZpA_33_60to226"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dashboard, Steering wheel, Radio? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the dashboard, steering wheel, and radio as requested. The dashboard is prominently featured with its controls and wood trim, the steering wheel is visible on the left side, and the radio unit with its buttons and display is centered in the frame. No conflicting elements are present that contradict the specified objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a interior, steering wheel, and radio as described. The dashboard is visible visible at visible various and display trim, the steering wheel is in with the left side, and the radio is is a display and display is on in the frame. The additional elements are present.\"\n would the description objects.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a25c5e14eea07418e5beb464c5900f1b0e186088
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe438004a47f0c0781c95db75db65e38ddbb270d9c7392e1791132c3f13905b5
+size 129333
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..4266ff798a15dcb8124761ecf323057013f4d1be
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men sitting at a table in a casual setting, possibly a restaurant or a cafe. The man on the left is wearing a blue shirt and is looking at the man on the right, who is wearing a yellow shirt and glasses. The man on the right is holding a fork and appears to be eating or about to eat. On the table, there are three white cups and a plate with food. The background features red chairs and a white table, suggesting a modern and clean interior design. The style of the video is candid and informal, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "kRyFzx7PCVM_12_153to304"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men sitting at a table; one in a blue shirt, the other in a yellow shirt with glasses. There are three white cups and a plate with food on the table.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men sitting at a table, one wearing a blue checkered shirt and the other wearing a yellow shirt with glasses, which matches the description. There are three white cups and a plate with food on the table, also matching the description. The only deviation is the appearance of a woman in a green dress in the last frame, but this does not contradict the core description, as the primary focus remains on the two men and the objects on the table.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men sitting at a table, one wearing a blue shirtered shirt and the other wearing a yellow shirt with glasses. which matches the description. There are three white cups on a plate with food on the table, which align the description. The setting additional is the presence of a red in the red shirt in the background frame, but this does not contradict the core description and as the question focus remains on the two men at the table on the table.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f946ec9154a6734a70f22727158d1145694ad598
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:750d7f35d8af0aa3b91133adfb0ee9e4b8ce10ebb1681cbb8c8b20614c2ec347
+size 114774
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..50185013e98aeef196264252c03b0bf900916098
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a dessert being enjoyed. The dessert is served in a white bowl with a textured pattern, placed on a clear glass plate. The dessert appears to be a creamy, white substance, possibly ice cream or whipped cream, topped with a sprinkle of powdered sugar. A spoon is being used to scoop the dessert, indicating that it is being eaten. The spoon is silver and has a small amount of the dessert on it. The background is blurred, focusing the viewer's attention on the dessert and the act of eating it. The style of the video is simple and straightforward, with a focus on the dessert and the act of eating it. The video does not contain any text or additional elements."
+    ],
+    "video_ids": [
+        "o9_4jq4X5CY_30_274to451"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dessert (white, creamy substance), spoon (silver, with a small amount of dessert on it), powdered sugar sprinkle? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dessert in a white ramekin, which appears creamy and is topped with powdered sugar. A silver spoon is seen scooping the dessert, with a small amount of it on the spoon. All elements described in the condition are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white that a white bowlamekin, which appears to and white being with what sugar. There silver spoon is present withoping the dessert, and a small amount of the on the spoon. The elements match in the condition are present and match with the video content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a4d395f9c4e5138723117bbbe40b3f0e8252c92a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83046f59fee4e31d5a6b60c49ca3ace9554d2de383f17b84bf128c625e9f05f3
+size 149898
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..8f57f8a6d041dd3a4878f08d093a4250c0df8673
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a yellow car with its hood open, revealing the engine. The engine is red and black, with various components visible, including a battery, a red air filter, and a black engine cover. The car is parked on a grassy area, and the hood is propped up with a black prop. The car appears to be a sports car, and the engine is well-maintained. The video is a close-up shot of the car's engine, focusing on the details of the engine components. The style of the video is straightforward and informative, providing a clear view of the car's engine."
+    ],
+    "video_ids": [
+        "B4iz-VVap1w_39_0to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Yellow car with a red and black engine, including a battery, a red air filter, and a black engine cover.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a yellow car with its hood open, revealing an engine that is predominantly red and black. The engine cover is black, and there is a visible battery and red air filter (or similar red component) in the engine bay. These elements match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a yellow car with a hood open, revealing a engine that is predominantly red and black. The engine cover is black, and there is a red red and a air filter,which similar component component) within the engine bay. The elements match the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..84fa8eb970347b5dae2c46dc805433ae56f53e77
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3923b4e3baca8c6e534b467f8a6506c199a7b333eada9fbe30accc5a0d754d39
+size 222173
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..6677905d59d183036c404dc65d9503548692ff29
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a demonstration of a CD burner in action. The style of the video is informative and technical, showcasing the device's functionality. The CD burner is shown in three stages: first, it is closed and appears to be a part of a larger device, possibly a multi-track recorder; second, it is partially open, revealing the internal components; and finally, it is fully open, with a CD inserted and being burned. The video likely includes narration or on-screen text explaining the process of burning a CD using the device. The focus is on the CD burner itself, with no additional elements or distractions in the background. The video is likely intended for an audience interested in technology or music production."
+    ],
+    "video_ids": [
+        "z0J5NPblTI4_16_20to174"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: CD burner, CD? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a CD burner feature with a CD partially ejected from the device, matching the description. The text overlay also explicitly labels it as 'CD Burner', confirming the object's presence and function.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a CD burner with, a CD being inserted and the device. which the description of The presence on and indicates mentions the as aCD burnerer', confirming the presence's identity. function.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..80c6f8929e608acf6af0d656d6b3981dad23a3a1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:465b167a123a79d8366a30e7728a0fb167b9ab6e2371e7d4c5fc0b562f78ad51
+size 199880
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..8d394f68cae25f15c4e9567da5ea89ebffdf1e20
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a silver sports car on display at a car show. The car is sleek and shiny, with a large front grille and a distinctive side vent. The car is parked on a gray carpet, and the background features a white wall with logos and text. The car is the main focus of the video, and it is presented in a way that highlights its design and features. The lighting in the video is bright, emphasizing the car's metallic finish and the reflections on its surface. The overall style of the video is sleek and modern, with a focus on the car's design and the car show setting."
+    ],
+    "video_ids": [
+        "nlEryNOR-l8_9_0to169"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A silver sports car with a large front grille and a distinctive side vent.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a silver sports car with a prominent front grille and a distinctive side vent, matching the description. The car's design features, including the grille and side vent, are clearly visible and align with the specified characteristics.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts a silver sports car with a sleek front grille and a side side vent, which the description provided The car's sleek features, such the grille and side vent, are clearly visible and align with the given characteristics.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2e64a92c4dd4cac3d09d736a02c7298a3ef73985
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:32b6c577a047a83bd6a339f153f761fe1d258426431137ea1f5594c6fb9354a2
+size 112950
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..91082deb3ae1896ef9c80d61a7bee77ba198998a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the process of cooking a hamburger patty on a stovetop. The first frame shows a clean, empty frying pan on the stove. In the second frame, a single hamburger patty is placed in the center of the pan. The third frame shows the hamburger patty being flipped with a spatula, indicating that it is being cooked on both sides. The style of the video is straightforward and focuses on the cooking process, with no additional elements or distractions. The video is likely intended for instructional purposes, such as teaching someone how to cook a hamburger patty."
+    ],
+    "video_ids": [
+        "hUtOjLsj2C0_13_0to151"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A single hamburger patty, a spatula? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a single hamburger patty cooking in a frying pan, with a metal spatula resting beside it. These are the only two objects mentioned in the condition, and they are both present and clearly visible throughout the video. No conflicting elements are present that contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a single hamburger patty in in a pan pan. which a spat spatula being on it. The elements the primary objects objects mentioned in the ', and they are both present in correctly visible in the video. The additional elements are introduced.\"\n would the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f5ec017ad4a09149e7295f195270d17dd9574764
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1203-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30330118e3edbebba7a6b76a381a7869e03cdf5a95d5b1531f28dc8244ca9b47
+size 123640
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..1fb0334ee7e40c9f756318b1f8d6d4dae3a3720e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with glasses, wearing a red shirt, sitting in front of a blue wall. The wall is adorned with various items, including pictures, a guitar, and a sports ball. The man appears to be engaged in a conversation or presentation, as he is looking directly at the camera. The overall style of the video suggests a casual and relaxed atmosphere, with the man as the main focus. The background elements add a personal touch to the setting, indicating that the video might be taking place in a home or a casual workspace."
+    ],
+    "video_ids": [
+        "88quO1OP6Gk_2_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with glasses wearing a red shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing glasses and a red shirt. He is the central subject throughout the frames, and there are no visual elements that contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing glasses and a red shirt. The is the central figure of the frames, and there are no additional elements that contradict the description.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..74e7a85273dbc8e751687d8bd307a567b3b29cc0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8dae5a8327a18fe2f593c63255eef3b790c20e0e565103f9afae3034bbf97b8b
+size 166485
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..6588e1c819fd7760f3c99f90d5c0bba0142c4f36
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures an intense moment in a football game. The main focus is on a player in a white jersey, who is in possession of the football. He is being tackled by two players in black jerseys, one of whom is wearing a yellow and black striped jersey. The player in white is trying to maintain his balance and protect the ball, while the two players in black are attempting to bring him down. The action takes place on a football field, with other players visible in the background. The style of the video is dynamic and action-packed, capturing the high-energy nature of the sport."
+    ],
+    "video_ids": [
+        "o2qlntM3qpo_63_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Player in a white jersey, two players in black jerseys, one of whom is wearing a yellow and black striped jersey.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a white jersey (Baltimore Ravens) holding the football, being tackled by two players in black jerseys. One of the tackling players is wearing a yellow and black striped jersey (Pittsburgh Steelers), matching the description. The scene is consistent with an American football game, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a football in a white jersey andlikelyears Ravens) being the football, and surrounded by two players in black jerseys ( One of the black players is wearing a yellow and black striped jersey,Pittsburgh Steelers). which the description. The presence is set with the American football game, and the additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..51a5c59ec56b3b8c56dcdeb2c6d313e75a1613f2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5a62944f2ab3922693541a95884d5bc0fdc121a20d26af330562d831b82a878
+size 268056
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..3a597d09cf8a3b9f9ae9df70ace9a0c48f49d851
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man wearing a headset and sitting in front of a computer monitor. The man appears to be engaged in a conversation or possibly playing a video game. The headset is black and has a microphone attached to it. The computer monitor is large and takes up a significant portion of the frame. The man is wearing a red shirt and is looking directly at the camera. The background is blurred, but it appears to be an indoor setting with a lamp and a shelf. The style of the video is realistic and it seems to be a still image rather than a moving one."
+    ],
+    "video_ids": [
+        "KtgeS2hhV9U_42_0to128"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man, headset, microphone, computer monitor, red shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a headset with a microphone, seated in front of a computer monitor, and dressed in a red shirt. All specified objects are present and accurately represented in the scene without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing a headset with a microphone, sitting in front of a computer monitor. and wearing in a red shirt. The the elements are present and match depicted in the video.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ea1e41d57a9157a7e304c10a72128e36e5339a33
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:56ddf5ab866ce16b8fffe2e9486861dfa5a3d842500c61c85a990ebd2be04a5f
+size 52958
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..54ad804a9af59705ba86cf23a15c524595de92f3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman named Jay Parry, who is the Super Bowl Host Committee Chair. She is seated at a desk with a football on it, suggesting a connection to the Super Bowl event. The setting appears to be an office or a newsroom, as indicated by the presence of a news ticker at the bottom of the screen. The news ticker displays the time as 8:47 and the temperature as 99 degrees, which could imply that the video was taken during a warm season or in a warm climate. The woman is dressed in a blue denim jacket, which adds a casual yet professional touch to her appearance. The overall style of the video seems to be a news segment or an interview, with a focus on the woman and her role in the Super Bowl event."
+    ],
+    "video_ids": [
+        "EBD83SfY9fU_2_42to173"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman named Jay Parry, a desk with a football on it, a news ticker? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman identified as Jay Parry, who is the Super Bowl Host Committee Chair, sitting at a desk with a football on it. A news ticker is also visible at the bottom of the screen, displaying information about Super Bowl preparations in Phoenix. All core elements described are present in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman sitting as Jay Parry, sitting is seated main Bowl MVP.'s. sitting at a desk with a football on it. There news ticker is visible visible at the bottom of the screen, displaying various related the Bowl L. Russia. The elements elements of in present in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..81702a5b7495ab8d64dd0b1b5dbaef7c707fe63c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff56643ab5f48909621cba3e106fc7b4b833c3f61647d102a4cc1bece43f94f6
+size 161073
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..bf2e4b0e403c72a79de14f433752822e9da196f4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a car engine, which is the main focus of the image. The engine is silver and black, with various parts and components visible. The engine is mounted on a red stand, which is positioned in the center of the image. The background is a plain white wall, which provides a stark contrast to the engine. The engine is shown in three different positions, each time revealing more of its internal components. The style of the video is informative, with a focus on the details of the engine. The camera angle is close-up, allowing for a detailed view of the engine's components. The video does not contain any text or additional graphics. The overall impression is that of a technical demonstration or educational video about car engines."
+    ],
+    "video_ids": [
+        "DKAqi03JnUE_8_127to248"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A silver and black car engine with visible parts and components, a red stand, and a plain white wall.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a silver and black car engine with visible parts and components, mounted on a red stand, against a plain white wall. All elements described in the condition are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a silver and black car engine with visible parts and components, which on a red stand. and a plain white wall. The elements of in the condition are present in match with the video content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..744e0368ab2627640157b390114bb894cab60ab8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:524ba20b55ba6619c4618730694b85d82c2681a65a6675bf71d185f1dfe5a060
+size 275811
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..b8f50dd84d4af05c7cb3aad2be6198afd893a889
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two men are preparing food at a food stall. The man on the left, wearing a white shirt with a rainbow logo, is standing behind a counter filled with various bowls and containers. He is looking at the man on the right, who is wearing a blue shirt with an orange stripe. The man on the right is using a knife to cut a piece of meat, which is placed on a white plate. The food stall is located in a busy area, with other people visible in the background. The atmosphere is bustling and lively, with the focus on the two men preparing the food."
+    ],
+    "video_ids": [
+        "CUDs9j_GVz8_19_0to192"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, a counter, a knife, a piece of meat, a white plate? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men at a counter, one cutting meat with a knife on a white plate. The core elements described are present: two men, a counter, a knife, a piece of meat, and a white plate. Additional elements like bowls, other people, and shelving do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men in a counter where with of a on a knife on a white plate. The presence elements of\u2014 present: two men, a counter, a knife, a piece of meat, and a white plate. The elements like bowls and a kitchen, and aving in not contradict the core and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9dce2952e02baecc21f8d80778e5c8df1f16fe2b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77b1479a0eebbe86f3ea062370256d9901029478190047afabd8e9eedc31b60e
+size 214353
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..782bb0d350c51ef3a42bb6eb4f9a8af546b9586c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and glasses, wearing a black cap and a black t-shirt. He is seated in the back of a car, with the car door open. The man appears to be in motion, as suggested by the blurred background. The man's facial expression is one of surprise or shock, with his mouth open and eyes wide. The style of the video is candid and informal, capturing a spontaneous moment in the man's life. The focus is on the man's face and expression, with the background serving to provide context and setting."
+    ],
+    "video_ids": [
+        "NdbFwQB1q74_24_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and glasses, wearing a black cap and a black t-shirt, a car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man with a beard and glasses, wearing a black cap and a black t-shirt, seated inside a car. The interior of the car, including the seats and windows, is visible, confirming the setting. The man's appearance and attire match the description, and there are no conflicting elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man with a beard and glasses, wearing a black cap and a black t-shirt. sitting inside a car. The presence of the car is including the window and windows, is visible, confirming the presence. The man's attire and attire match the description provided and there are no additional elements in would the given description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..434611173b54f2d407c445575a0399b5fb5416f9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea2d1e924d5a5c9e1a2ae94e9294ff31a19be4c363b4f67b2653ac3675d7de8e
+size 177732
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..a8e8bda78a54e034fcac9582bfcabafb2c6b6704
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up view of the interior of a car, focusing on the driver's side. The car has a sleek design with black leather seats and a red stitching detail. The steering wheel is black with silver accents and the Mercedes-Benz logo prominently displayed. The dashboard is also black with silver accents and features a variety of buttons and dials. The car's interior is well-lit, highlighting the details of the design and the quality of the materials used. The video is likely a promotional or review video for the car, showcasing its interior design and features."
+    ],
+    "video_ids": [
+        "QQS11qUWcTI_6_0to119"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Black leather seats, red stitching, black steering wheel with silver accents and Mercedes-Benz logo, black dashboard with silver accents and various buttons/dials.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows black leather seats with red stitching, a black steering wheel featuring silver accents and the Mercedes-Benz logo, and a black dashboard with silver accents and various buttons/dials. All elements described in the condition are present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a leather seats with red stitching, a black steering wheel with silver accents and the Mercedes-Benz logo, and a black dashboard with silver accents and various buttons andials. The elements in in the condition are present and match depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9a32a53fc04d77f3b45d02fbfaab8b8901f215cb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1204-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a3e0b4e71c20df511e09c781f8aafa80cb2ea8cb768fcb0deb617d19d09cc4d2
+size 140437
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..447004e5016cb5cf46f1a8004a3519008a52712d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young couple is seen embracing each other in a warm and affectionate hug. They are standing in front of a brick building, which is adorned with palm trees. The man is wearing a blue shirt and a backpack, while the woman is dressed in a white shirt and a pink backpack. The couple's happiness is evident in their smiles and the way they hold each other. The video captures a moment of love and affection between the two, set against the backdrop of a beautiful day."
+    ],
+    "video_ids": [
+        "3dupaeI-7ZQ_14_132to290"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young couple embracing, a man in a blue shirt with a backpack, a woman in a white shirt with a pink backpack.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young couple embracing, with a man in a blue shirt and a backpack, and a woman in a white shirt with a pink backpack. Their actions and attire match the description, and additional elements (like another person walking by) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a young couple embracing, with the man wearing a blue shirt and a woman, and a woman in a white shirt with a pink backpack. The attire and attire match the description provided and there elements likelike the person in in in do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0809c7b153c710ebde9e817230df8c93c471424f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c9683a730aee7a2c2e04b769f717227e41da7c5124c940ee1da7e879b1139081
+size 203130
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..1616bf63421545dcb65235b0cb4b211220a2b494
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and a bald head, wearing a striped shirt and an orange tie. He is holding a sign that reads \"Olive Garden\". The man appears to be speaking, possibly presenting or promoting the restaurant. The style of the video is straightforward and promotional, with a focus on the man and the sign he is holding. The background is a simple, uncluttered green wall, which puts the emphasis on the man and the sign. The video is likely intended for advertising or informational purposes, highlighting the name of the restaurant and the man's association with it."
+    ],
+    "video_ids": [
+        "XMNHFqS1Fug_44_0to144"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: man with a beard and a bald head, wearing a striped shirt and an orange tie, sign reading 'Olive Garden'? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a bald head and a beard, wearing a striped shirt and an orange tie, holding a sign that reads 'Olive Garden'. All elements of the description are accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard head and a beard, wearing a striped shirt and an orange tie. which a sign that reads 'Olive Garden'. The the of the description are present represented in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..36ab8ea37ff09d564b04ef8351f3a85792b9f8f8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d8aee5859723122642383ee413fbfe3367889b008994ee75c46a686e6674306e
+size 337727
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..dd93e1f877c0e9e13946baa6ac5f54fc7b537c97
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment about kidney failure. The style of the video is informative and educational. The first frame shows a man with glasses, presumably the news anchor, speaking about the topic. The second frame shows a close-up of a medical book with the title \"KIDNEY FAILURE\" written on it, indicating the subject of the news segment. The third frame shows a hand holding a stethoscope, suggesting a medical context. The overall tone of the video is serious and informative, aimed at educating viewers about the topic of kidney failure."
+    ],
+    "video_ids": [
+        "GljqJLrJy_U_12_28to187"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man with glasses, medical book, hand holding a stethoscope? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with glasses speaking, a medical book titled 'KIDNEY FAILURE' being held by a hand, and a stethoscope visible on the table. These elements match the 'Object(s)' condition described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a man with glasses,, which hand book with 'KINDNEY R' is written, a hand, and a handethoscope being in the hand. The elements collectively the descriptionObject(s)' condition provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..93b2cf98cc186b006ebd7ca5c708d56851a146a0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a60fcbddede19d1f5d9560566edf1cf5cc164139b6a81bc48fdaee548a488ad
+size 198926
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..6d5548b191c532874a0c66fe69e99ddf20db1491
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit and tie, wearing glasses, standing in an office environment. The man is speaking, and his expression is serious. The office setting includes a blurred background with a desk and a computer monitor. The lighting in the scene is bright, and the overall style of the video is professional and polished."
+    ],
+    "video_ids": [
+        "4T8R1GJ5zdI_3_0to116"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, wearing glasses, standing and speaking.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a suit and tie, glasses, standing, and speaking. The background is blurred but does not contradict the description. All core elements of the condition are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit and tie, and, and in and speaking. The setting appears an but appears not contradict the description. The elements elements of the description are met represented.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7d7035f72733347e0c528a38320939a5efe7c296
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5fbae11994895c478778ddb0d4409f630cac17b068177a437e1b23b7d99841a
+size 92435
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..3cbf85c8ba0fbd68dfcd708e8a1fce75efa36226
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a character from a video game, standing in a dark, stone-walled room. The character is dressed in a red jacket and holds a sword in their right hand. The sword is pointed upwards, and the character's face is illuminated by a bright light. The character's expression is serious, and they appear to be ready for battle. The room has a medieval feel, with stone walls and a large fireplace in the background. The overall style of the video is realistic, with detailed textures and lighting effects that create a sense of depth and realism."
+    ],
+    "video_ids": [
+        "R-GfY9mA6Lg_39_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A character from a video game, dressed in a red jacket, holding a sword in their right hand, and a large fireplace in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a character from a video game (Dante from Devil May Cry), dressed in a red jacket, holding a sword in his right hand. In the background, there are flaming torches mounted on stone pillars, which can be interpreted as a large fireplace or fire feature. The scene matches the core description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a character from a video game dressedasante from Devil May Cry) dressed in a red jacket, holding a sword in their right hand. The the background, there is elements logses, on the walls, which can be interpreted as a large fireplace. a feature, The setting is the description description provided contradict.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4a8764ffff4707b166a1a0472b45369df9b142b9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:738ae30f5be542f1400f8a5a96eaf879c2d453b1a9d9c6cf12b93cdb7af096e6
+size 233791
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..01c2aa3639be9cdd7cdc3567177b27f154cd4086
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a bowl of soup being enjoyed. The soup is served in a black bowl with a speckled pattern, and a silver spoon is placed within it. The soup itself is a rich orange color, suggesting a hearty and flavorful broth. The spoon is partially submerged in the soup, indicating that it is being used to enjoy the dish. The bowl is placed on a wooden table, which adds a rustic touch to the scene. The overall style of the video is simple and straightforward, focusing on the act of enjoying a warm and comforting meal."
+    ],
+    "video_ids": [
+        "iD4nLXnJhs0_64_59to195"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Black bowl with speckled pattern, silver spoon, soup (rich orange color).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a black bowl with a speckled pattern, a silver spoon resting inside, and a rich orange-colored soup with visible ingredients. These elements match the description exactly, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a black bowl with a speckled pattern, a silver spoon, in it and a rich orange-colored soup. a texture, The elements match the description provided, fulfilling there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c96001cf4eb239e181ed9dc0ac49cc7e611201fa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee963c4b237f45c701120732342ebe108a013f5e3c8206e0a58b8dc1b0c8de86
+size 127115
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..c97cb0e02e83f96f041df27cc29fe4bd7a473ce4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a female character in a futuristic setting, possibly from a video game or animated series. She is dressed in a white and red outfit with medical patches and a helmet, suggesting she is a medic or healer. In the first frame, she is seen holding a large stack of medical supplies, which she then tosses into the air. In the second frame, she is seen waving her hand, possibly in a gesture of greeting or farewell. In the third frame, she is seen holding a large medical device, which she appears to be using to heal or revive someone. The background is a clear blue sky, indicating an outdoor setting. The style of the video is colorful and vibrant, with a focus on the character and her actions."
+    ],
+    "video_ids": [
+        "2Cghin8WHyM_129_0to117"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Female character, large stack of medical supplies, large medical device? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a female character dressed in medical gear, holding a large medical device resembling a syringe or IV pump. A large stack of medical supplies, including bandages and first aid kits, is also visible. These elements align with the 'Object(s)' condition described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a female character wearing in a attire, holding a large stack device and a defringe or a pump, She large stack of medical supplies is possibly boxesages and other aid kits, is also visible. The elements align with the descriptionObject(s)' condition,,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6525fd5ac1c43a0ffc78a404c1fd78174533b8ea
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:654fc3253a0cdc1a77de0b34b4f3ec5eece85e56ee79ccebd3f7cce18c5f1e6a
+size 328676
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..49935d8b5fdb6868e5870859def77746d1e72f01
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two men are seen cooking outdoors using a traditional method. They are using a large metal pot, which is placed over a fire made from logs. The men are dressed in traditional attire, suggesting that the video may be set in a rural or traditional setting. The fire is burning brightly, indicating that the cooking process is underway. The men are actively engaged in the cooking process, with one man stirring the pot and the other man tending to the fire. The overall style of the video is rustic and traditional, capturing a moment of communal cooking in a natural setting."
+    ],
+    "video_ids": [
+        "76V3FZX-REs_23_0to155"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, a large metal pot, logs for the fire.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men carrying and placing a large metal pot over a fire made of logs. The core elements described \u2014 two men, a large metal pot, and logs for the fire \u2014 are all present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two men cooking out stirring a large metal pot over a fire made of logs. The presence elements of in the men, a large metal pot, and logs for the fire \u2014 are all present and accurately depicted in the video.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..28ab9d941f6207091fa1fcd091ebbeb8879db7b2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1205-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2ec24aa5fadb8b868b49dd521f1c78dca3ff39f219c9fa136f1260bf8ab4c4ec
+size 321954
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..5cce4173f4484a23246627cd9eb505143b25b92e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a cooking tutorial featuring a woman in a pink tank top preparing a dish. She is standing at a kitchen counter with various ingredients and utensils. The counter has a wooden cutting board, a bowl filled with chopped fruits, a bottle of maple syrup, and a jar of chia seeds. The woman is holding a spoon and appears to be in the process of mixing the ingredients. The video includes a text overlay that reads \"CHIA SEEDS 2 to 3 tablespoons\" and the logo of the brand \"THRIVE\". The style of the video is instructional and focuses on the preparation of a healthy snack or meal."
+    ],
+    "video_ids": [
+        "KMw-pdEWYdk_30_0to190"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Woman in a pink tank top, wooden cutting board, bowl of chopped fruits, bottle of maple syrup, jar of chia seeds, spoon.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a pink tank top preparing a fruit bowl on a wooden cutting board. She adds chia seeds and maple syrup to the bowl of chopped fruits. The bottle of maple syrup and jar of chia seeds are visible on the counter, and although a spoon is not explicitly shown being used, the context implies it might be used for mixing or measuring. All core elements from the description are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a pink tank top preparing a dish dish. a wooden cutting board. There is chia seeds to maple syrup to the bowl, chopped fruits. The presence of maple syrup and the of chia seeds are visible on the cutting, and a the spoon is not explicitly mentioned in used in it context suggests that is be used for adding. serving the The the elements of the description are present and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..480245119186333bb67c8c5da6629bc31a1948df
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b94595b7d1b392a14f9a65c8894918cca0b18dba2e90deb77ccea1fd480dd73
+size 131689
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..1ab3cefd1cb11fe11be29801c613249f61603083
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a woman posing with three pink sports cars in a parking lot. The woman is wearing a red dress and is standing between the two middle cars, with her hands on her hips. The cars are parked in a row, and the woman is positioned in the center. The cars are pink with black accents, and they have the Chevrolet logo on the front. The parking lot is surrounded by palm trees, and there are buildings in the background. The sky is clear and blue, and the sun is shining. The style of the video is a combination of a car advertisement and a fashion shoot."
+    ],
+    "video_ids": [
+        "6OxDBAMDF4A_22_0to134"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, three pink sports cars with black accents and Chevrolet logos, palm trees, buildings.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a woman posing between three pink sports cars with black accents and Chevrolet logos. Palm trees and buildings are also visible in the background, matching the description. The video is a composite or overlay, but the core elements described are present and consistent.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a woman in between three pink sports cars with black accents and Chevrolet logos. The trees and buildings are also visible in the background, matching the description. The presence does a still image AI, as it elements elements are are present and accurately with}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ed2c7519e12347eaeffa8cf09841949588c9de0b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb34fe4e2ff2c2a4915934ff57fde9bf01124af6900eefb3e36107f56aa151ef
+size 206256
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..602963ee3f88b58ebdbe73366e6c22d5a8ac674b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a playful and imaginative mashup of animated characters from different movies and universes. It features a gold robot standing confidently in the center of a room filled with animated princesses. The princesses are dressed in colorful gowns and are engaged in various activities, such as dancing, playing, and conversing. The room is decorated with elegant furniture and a chandelier, creating a festive atmosphere. The robot's presence adds a touch of humor and surprise to the scene, as it contrasts with the traditional fairy-tale setting. The video is likely designed to entertain and amuse viewers by combining elements from different genres and franchises."
+    ],
+    "video_ids": [
+        "lJ-hbJDWazY_14_24to160"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A gold robot and several animated princesses engaged in various activities.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a prominent gold robot (C-3PO) and multiple animated princesses from Disney films, all engaged in various activities such as dancing, standing, and interacting. The scene is set in a luxurious room, and the characters are clearly visible and consistent with the described condition. No elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a gold gold robot standinglikely-3PO) standing several animated princesses in the's. engaging engaged in what activities such as dancing and interacting, and interacting with The setting is set in a room room with which the characters are dressed animated, distinct with the description elements. The additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..396bde4b538a5f6db8b0be1e1a76eeae33a62bfc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7313c08ff40dcce8f3c539bead6dfd42236b320fbcef866bbc30d6cc44d3d7f2
+size 232394
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..a2dd0acb97d28a8f78141bf7778d79c1dafeb36c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up of a black and green motorcycle helmet with a clear visor. The helmet is shown from the side, with the visor partially open, revealing the interior of the helmet. The helmet has a reflective surface, and the visor has a slight tint. The helmet is positioned against a blurred background, which suggests that the focus is on the helmet itself. The style of the video is a product demonstration, showcasing the features of the helmet. The video is likely intended for a motorcycle enthusiast audience, highlighting the design and functionality of the helmet."
+    ],
+    "video_ids": [
+        "B6BbQ0-2TYQ_13_0to175"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black and green motorcycle helmet with a clear visor, partially open to reveal the interior.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black and green motorcycle helmet with a clear visor, and the visor is partially open, revealing the interior. The description matches the visual content accurately, with no contradictory elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close and green motorcycle helmet with a clear visor. which the visor is partially open, revealing the interior of The helmet matches the visual content of, with no additional elements present}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..68cc601bc62ed99b457511a73856be497276317a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a2faf4f2aa3f94da03e82efb589c07d4a191b7ec7349168ad8b73c408aeb9633
+size 94732
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..9299e1775766276d1bbe261d1ed70ffdbf7c8e63
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the grandeur of the Las Vegas Ballpark, a modern baseball stadium located in the heart of Las Vegas. The first frame shows the stadium from a high angle, revealing its impressive architecture and the vibrant cityscape that surrounds it. The second frame offers a closer view of the stadium, highlighting its large, curved roof and the bright lights that illuminate the field. The third frame provides a ground-level perspective, showcasing the stadium's entrance and the bustling activity around it. The video is a testament to the city's love for baseball and the allure of Las Vegas's entertainment scene."
+    ],
+    "video_ids": [
+        "Jrn-AIIH4Oc_3_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Stadium, large curved roof, bright lights, stadium entrance, people? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing the Las Vegas Ballpark stadium with its large curved roof, bright lights illuminating the field and surrounding areas, and visible stadium entrance. Although people are not prominently featured, their presence is implied by the activity around the stadium and the lighting, which suggests occupancy or preparation for an event. The core elements described are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a following Vegas Raiderspark, with a large curved roof, bright lights,ating the entrance and entrance areas, and a stadium entrance. There there are not explicitly featured in the presence is implied by the activity around the stadium, the lighting that which suggests the. preparation for an event. The video elements of in present represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..493fd5642f43e2880a4d94b0e3a398cd45ca5c7c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b30bd64d4d6ea3424b63e3254a4dab4f986e946b3d83bbef67ec740b5eb93a92
+size 224469
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..5c89ea90c9488ff52374e920122e3f42335d8aa3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse of an artist's sketchbook being filled with watercolor paintings. The artist's tools, including a paintbrush, a ruler, and a palette with various colors, are visible on the table. The sketchbook is open to a page with several watercolor paintings, each depicting a different scene or subject. The style of the video is artistic and creative, showcasing the process of creating watercolor paintings. The artist's tools and the open sketchbook are the main focus of the video, with the watercolor paintings providing a visual representation of the artist's work. The video captures the beauty and creativity of watercolor painting, as well as the process of creating art."
+    ],
+    "video_ids": [
+        "mr6XJgSzGJs_7_85to306"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Artist's tools (paintbrush, ruler, palette), sketchbook with watercolor paintings? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows an open sketchbook with watercolor paintings, including a detailed illustration of a person with a cup and books, and the word 'April'. Artist's tools such as paintbrushes, a ruler, and a watercolor palette are visible on the wooden table around the sketchbook. These elements align perfectly with the described 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a artist sketchbook with watercolor paintings, which various variety landscape of a landscape. a green. a. which a presence 'LO'. The's tools such as abrushes and a ruler, and a palettecolor palette are also in the table surface. the sketchbook. The elements match with with the description 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7b22fed064313b67d6c685ac4631ac70ee1ca9d4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b50cc12fe6e67e6ce48582ec251cd55fd1e6f940bbf7f08801079d74c40a5278
+size 178610
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..886b6361b3565c80de7203aafa2a9652fa9349f2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a young man sitting in a classroom, wearing a brown beanie and a black and white scarf. He is looking to his left with a serious expression on his face. The classroom is filled with other students, some of whom are looking at him. The man is holding a book in his hands, suggesting that he is either reading or about to read. The overall atmosphere of the video is one of concentration and focus."
+    ],
+    "video_ids": [
+        "b9SQf30YHvI_236_115to278"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man, other students, a book.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a young man wearing a beanie and scarf, sitting among other students in what appears to be a classroom or lecture setting. He is holding and looking at a book, which aligns with the 'Object(s)' condition. Other students are visible in the background, and the book is prominently featured as an object in the foreground. There are no elements that contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a young man wearing a brownanie and scarf, holding at other students in what appears to be a classroom setting lecture setting. He is holding and reading at a book, which aligns with the descriptionObject(s)' condition. The students are also in the background, and the setting is a featured in the object in the foreground. The are no elements in contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..eb8575e1628a39634fa4c6b41b039fc6e63f4014
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21715db67cd3a7817e04feb9a95b0d1b7ba50e64a052893d035c856d2f58c543
+size 184595
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..e904a47bbb5f3eba02951a074dff09b599fffffe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a black and red vehicle, possibly a car or a truck, with a focus on its side. The vehicle has a black door with a red handle and a red trim. On the door, there are several stickers and decals, including one that reads \"RACELINE WHEELS\" and another that says \"FACTOR 55.\" The vehicle also has a red fender with a black trim. The style of the video is a straightforward, close-up shot of the vehicle, focusing on its details and accessories. The video does not show any movement or action, but rather provides a static view of the vehicle's side."
+    ],
+    "video_ids": [
+        "iKquSuq-Uts_75_70to243"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Black and red vehicle, black door, red handle, red trim, stickers/decal 'RACELINE WHEELS', stickers/decal 'FACTOR 55', red fender, black trim? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black and red vehicle with a black door, red trim, and a red handle. Stickers for 'RACELINE WHEELS' and 'FACTOR 55' are visible on the side. The red fender with black trim is also clearly shown. All core elements described are present in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black and red vehicle with a black door, red handle, and a red f. Thereickers/de 'RACELINE WHEELS' and 'FACTOR 55' are visible on the door of The red fender and a trim is also present seen. The elements elements of in present in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6d28f880ce5f7eec0c8605ce23c08826e08afa2c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1206-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d0bcc61980e432b383de330391c388d965b398046d07a395c80847763160e75
+size 102919
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..372dddc7be4b81f0b691c2b0b5d797c7d6418c76
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a vibrant and colorful collage of various fruits and vegetables, each frame showcasing a different type of produce. The first frame features a bowl of green olives, their glossy surface reflecting the light. The second frame displays a bunch of ripe red tomatoes, their bright color contrasting with the wooden surface they rest on. The third frame presents a half of an avocado, its creamy green flesh and dark brown pit visible. Each frame is marked with a green checkmark, indicating approval or selection. The style of the video is simple yet appealing, with a focus on the freshness and natural beauty of the fruits and vegetables."
+    ],
+    "video_ids": [
+        "-PLN65j2rk8_5_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Green olives, ripe red tomatoes, half an avocado? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing green olives in a bowl, ripe red tomatoes (one whole and one cut in half), and half an avocado. These items are prominently displayed and match the description exactly.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video contains fulfills the 'Object(s)' condition as showcasing showing green olives, a bowl, ripe red tomatoes,which of and one sliced in half), and a an avocado. The objects are the displayed and match the description provided.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2cb0615748e66d7fee8edfc8a31b07e19a0f3985
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:94f577c07e2b568ed9648b240a480b4ea6af24db2b03e1e6d75c549dae727c72
+size 59063
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..5c9295502b5cd56a96cb02457db63911a3187475
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman with long brown hair is seen in a brightly lit room with a white wall and a red plant in the background. She is wearing a blue floral blouse and has a surprised expression on her face. Her eyes are wide open and her mouth is slightly open as if she is in the middle of a conversation or reacting to something. The overall style of the video is casual and candid, capturing a moment of surprise or interest in the woman's expression."
+    ],
+    "video_ids": [
+        "6BO2xOjooKY_4_20to197"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with long brown hair, wearing a blue floral blouse.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with long brown hair wearing a blue floral blouse, which matches the description. The background elements, such as the plant and wall decor, do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with long brown hair wearing a blue floral blouse. which matches the description provided The background includes, such as the red, the,, do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..868d4b41f9837767e5acae45b37f93135273e5b0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07503c6dfaf98e4485b325d2addc8a452a3c5866450217b748ef9700b39735b6
+size 240796
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..3b7f22b19d2f05723207afaa2d09d15cd09610d9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a bakery display case filled with a variety of donuts. The donuts are arranged in three different trays, each containing a different type of donut. The first tray contains chocolate donuts, the second tray contains glazed donuts, and the third tray contains chocolate donuts. The donuts are neatly arranged in rows, and each tray has a label indicating the type of donut. The bakery display case is clean and well-lit, highlighting the colors and textures of the donuts. The video is a simple, straightforward depiction of a bakery display case, focusing on the variety and arrangement of the donuts."
+    ],
+    "video_ids": [
+        "LuMeLU35kUQ_181_112to243"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three trays of donuts, labels indicating the type of donuts.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three distinct trays of baked goods, which are labeled as 'EARL GRAY' and other types (implied by the presence of different appearances and labels). Although the items are more accurately described as cookies rather than donuts, the core requirement of having labeled trays of baked goods is fulfilled. The labels are clearly visible, indicating the type of baked goods, which matches the description's intent.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three trays trays of don goods that which appear don with dondonL GREAY' and ' flavors,thoughplied by the different of different designs and colors). The the video are not likely described as don or than donuts, the video condition of having three trays of baked goods is met. The trays are visible visible, indicating the type of baked goods in which align the description.\"\n condition.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..39c90160801692d6da75372688824c6a5813225f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ca5c907c935c34f1619afb9d3ef29b402982ea83c94bb48cbbb8cce0b2cdb7c
+size 177911
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..5a361dada08e13e4960f6289a5b8cbf846bbe96b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the growth and transformation of a plant over time. The first frame shows a small, green leaf sprouting from a branch. In the second frame, the leaf has grown larger and is now a vibrant shade of green. By the third frame, the leaf has reached its full size and is a deep, rich green color. The plant's growth is evident in the progression from the first to the third frame, showcasing the plant's development and vitality. The style of the video is a close-up, time-lapse shot that focuses on the plant's growth, providing a detailed view of the plant's transformation over time."
+    ],
+    "video_ids": [
+        "tFu8U9ve9XI_29_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A leaf? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features leaves, specifically those of a plant with glossy, green foliage. The focus remains on the leaves throughout the clip, fulfilling the 'Object(s): A leaf' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video primarily features a, starting a of a heart with heart, heart,. The sequence is on the leaves throughout the sequence, with the 'Object(s)' A leaf' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..20d466bcaeb3adc6721386285b95677dd927365d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90884a67dcf23f8a81291361e428cb9ed0cf619e89cfd6f5e21a7ed78e913734
+size 176276
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..5163ed24ad230ce74f6f1326b2400f88be8646cc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a DIY water filtration system set up on a wooden table against a dark background. The system consists of a clear plastic bottle, repurposed into a filter, connected to a smaller rectangular aquarium by a flexible tube. The bottle is filled with water and contains a red stone at its base, while the aquarium holds a small fish, some aquatic plants, and colorful pebbles at the bottom. A hand appears in the frame, adjusting a yellow valve on the tube, which controls the flow of water from the bottle to the aquarium. As the valve is adjusted, bubbles can be seen rising in the aquarium, indicating the movement of water through the filtration process. The setup demonstrates an innovative way to maintain a healthy aquatic environment using simple materials. The video captures the interaction between the hand and the valve, emphasizing the functionality of the filtration system."
+    ],
+    "video_ids": [
+        "1c2cd3bfecfdd1ff8af3e73be8d4b23a125db11335bfed482625aaa6a3481ff9"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Clear plastic bottle, smaller rectangular aquarium, flexible tube, red stone, small fish, aquatic plants, colorful pebbles, hand, yellow valve.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows all the specified objects: a clear plastic bottle (modified as an air pump), a smaller rectangular aquarium with water, a flexible tube connecting the bottle to the aquarium, a red stone, small fish swimming, aquatic plants, colorful pebbles at the bottom, a hand adjusting the yellow valve, and the yellow valve itself. No conflicting elements are present that contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a the elements objects: a clear plastic bottle,used to a aquarium pump), a smaller rectangular aquarium, aquatic, a flexible tube connecting the bottle to the aquarium, a red stone, small fish, in aquatic plants, and pebbles, the bottom, and hand manipulating the yellow valve, and the yellow valve itself. The additional elements are present, would the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fb054d1d960de27ba312b36c1b56d55f4191273d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:58edf5d3efb081d97fe5c2aa381bd52b92c423f53a197c5dc911719eb925b1f4
+size 138807
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..5a85f152e975936021245d4722fb844184cde034
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen taking a selfie with his phone. He is wearing a white vest and a black mask. The man is standing in front of a tree, which is located on a street. The street is lined with buildings, and there are trees in the background. The man is holding the phone in his right hand, and he is looking at the screen. The phone is a blue color. The man appears to be in a good mood, as he is smiling while taking the selfie. The video captures a moment of everyday life, with the man enjoying his time outdoors."
+    ],
+    "video_ids": [
+        "3VsoJTYLgJA_3_0to192"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a tree, buildings, and a phone? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a mask and holding a phone, standing in front of a tree with buildings visible in the background. All specified objects are present and accurately represented in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man holding a white and a a phone, which in front of a tree and buildings in in the background. The the elements ( present and the depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..62aedb90aba9f1ee3144f37ead891215b8a29951
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e8a882dcd3fe256e0a362be6e462d518435a9c35307c94cb63e01413336ca40f
+size 168419
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..198056e96f4532d9596c86cbe4d537c845c68740
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a workshop filled with bicycles. He is wearing a black t-shirt with a logo on it. The man is gesturing with his right hand, possibly explaining something or pointing at something. The workshop is filled with various bicycles, some of which are stacked on top of each other. The man appears to be in the middle of a conversation or demonstration. The style of the video is casual and informative, with a focus on the man and his surroundings."
+    ],
+    "video_ids": [
+        "JRo7Gdo5nMc_5_0to110"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, various bicycles? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man speaking, and multiple bicycles are visible in the background, including a teal bicycle and several green frames. The core elements described in the condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man in in and there bicycles are visible in the background. fulfilling a yellow bicycle, a others bicycles. The setting elements of in the question are present, match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8cfd57d8185dfffb0cc5d50a18e0200261182ff4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b40998add074de611c1d13c5d43a4a48edc0346797757c6cbfabd5d99b37f2a
+size 282115
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..f7a75356db482806ae41d8659fa650b30fff06dc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a classic car, showcasing its vintage design and features. The car's dashboard is adorned with a variety of gauges and dials, each serving a specific purpose in monitoring the vehicle's performance. The steering wheel, a prominent feature in the center of the dashboard, is made of wood and metal, adding a touch of elegance and sophistication to the overall design. The car's interior is bathed in sunlight, highlighting the details of the dashboard and steering wheel. The video is a testament to the craftsmanship and attention to detail that went into designing this classic car."
+    ],
+    "video_ids": [
+        "x7skkkY-xq4_50_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dashboard gauges and dials, steering wheel? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the dashboard gauges and dials, along with the steering wheel, which are the primary objects mentioned in the condition. The camera focuses on these elements, and they are prominently displayed throughout the video, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a dashboard gauges and dials, as with the steering wheel, which are the core elements mentioned in the description. The image angle on these elements, providing there are prominently displayed in the frames. fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3d01a2134a02d4f6e0c82173347e08fa9d54796d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1207-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b83aedde3132c26ccde1f0b894cadbcf3e5f4d8cbea41be61dfe47a55487a327
+size 176924
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..f61d9ae6e9087d6b460a815c23d5f1c6d13e2fad
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a store display featuring a variety of white ceramic containers with black lettering, each labeled with a different vegetable or herb name. The containers are arranged on wooden tables and stools, with some placed on top of wicker baskets. The style of the video is straightforward and commercial, showcasing the items for sale in a retail setting. The focus is on the containers and their labels, with no additional action or movement. The lighting is bright and even, highlighting the white ceramic and the black lettering. The background is neutral and does not distract from the items on display. The video is likely intended for promotional purposes, to showcase the store's selection of kitchenware."
+    ],
+    "video_ids": [
+        "A9WwVHDOVYE_54_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: White ceramic containers with black lettering, labeled with vegetable or herb names.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple white ceramic containers with black lettering, clearly labeled with vegetable or herb names such as 'Onion', 'Potatoes', and 'Vegetables'. These items are prominently displayed on wooden tables and stools in a retail setting, matching the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows white white ceramic containers with black lettering, which labeled with names or herb names such as 'bion', 'Bato', ' 'Gargetables'. The containers are neatly displayed on wooden stands, are, a well or, which the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ea56f8953b768ffef941f825a4078fe52b5c2591
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d357580a1e040b05196e31f35a755f2868ec3aae30b127ab0b8b4c544f3347f0
+size 103518
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..6e46f0ff360eb5eca5125aea9cb4c109dd8af50c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring Lt. Joe Mendoza from the LA County Sheriff's Dept. He is standing in front of a chain-link fence, wearing a suit and glasses. The style of the video is a standard news report with a blue banner at the bottom displaying the name of the news channel and the time. The focus is on Lt. Mendoza, who appears to be speaking to the camera. The background is nondescript, with the chain-link fence and a fence post visible. The lighting is bright, suggesting it might be daytime. The overall tone of the video is serious, as is typical for news reports."
+    ],
+    "video_ids": [
+        "5gAlQQAGL-k_8_0to169"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Lt. Joe Mendoza (standing in front of the chain-link fence, wearing a suit and glasses), a blue banner at the bottom of the screen displaying the news channel's name and time.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows Lt. Joe Mendoza standing in front of a chain-link fence, wearing a suit and glasses, as described. Additionally, a blue banner at the bottom of the screen displays the news channel's name (KCAL 9) and time (12:02), matching the specified condition. The presence of other individuals and microphones does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a. Joe Mendoza standing in front of a chain-link fence, wearing a suit and glasses, which described. The, there blue banner at the bottom of the screen displays the news channel's name andthoughTL 9) and time (10:00), which the description condition.\"\n The presence of the elements in thephones in not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5727e54d5ab2b4bb72b0bba2a134f335a605d242
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:74e66d6feec5934d072cc8913241f63d804c0789f29d056739e32d356ba92ef3
+size 205857
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..873f94558fa01b9efa0a6d4a97e3ff3031d1c47b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a documentary-style film featuring a woman in a traditional Asian hat and black dress, standing amidst a lush green coffee plantation. She is holding a basket filled with ripe coffee beans. The scene captures the essence of coffee farming, showcasing the vibrant green leaves of the coffee plants and the rich, dark color of the coffee beans. The woman's attire and the setting suggest a connection to traditional farming practices, emphasizing the hands-on nature of coffee cultivation. The overall atmosphere is serene and peaceful, with the focus on the woman and her interaction with the coffee plants."
+    ],
+    "video_ids": [
+        "G3MdoPJseWY_27_0to184"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Woman in a traditional Asian hat and black dress, holding a basket filled with ripe coffee beans.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a traditional Asian conical hat and a black dress with red embroidery, standing in a coffee plantation. She is holding a basket filled with ripe coffee beans, which matches the core description. The background and setting are consistent with a coffee farm, and there are no elements that contradict the specified object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a woman wearing a traditional Asian hatical hat and a black dress, p coffee. holding in a lush plantation. She is holding a basket filled with ripe coffee beans. which align the description description. The setting is the are consistent with a coffee farm, and the are no additional that contradict the description object condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0f433ac9fc7eedbca636cbb775c92b924c21ac95
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a67585bb43a547bb0827fb7e382926a00d9cbcfa931ff104da43613510fc86cd
+size 195475
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..d8f32d61effe3bf9556860569b33838f399f86e3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a small, shaggy-haired dog with a light brown coat and a black nose. The dog is sitting on a white, fluffy surface, possibly a bed or a couch. The dog's ears are perked up, and it is looking directly at the camera with a curious expression. The background is blurred, but it appears to be a room with a patterned pillow or blanket. The style of the video is a close-up, portrait-style shot that focuses on the dog's face and upper body. The lighting is soft and natural, suggesting an indoor setting with natural light. The overall mood of the video is calm and peaceful, with the dog appearing relaxed and comfortable in its environment."
+    ],
+    "video_ids": [
+        "pcVKcjXrUAg_9_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small, shaggy-haired dog with a light brown coat and a black nose.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small, shaggy-haired dog with a light brown coat and a black nose, which matches the description. The dog is the central focus, and while there are additional elements like a background pillow and an Instagram handle overlay, they do not contradict the core description of the dog.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a small dog shaggy-haired dog with a light brown coat and a black nose, which matches the description provided The dog's lying main focus, and there there are some elements like a white and and a object-like,, these do not contradict the core description of the dog.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a5a171b91d9426921ab46ca4a6855fcb4abf4179
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5e303f2b4708301664740074c3b4be38ee26f1344b26f1da33846b5940c49a3
+size 129107
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..f808dc3772b58126646e075d62726f9e174540a9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a river flowing through a rural landscape. The river, with its calm and clear waters, is the central focus of the video. It is surrounded by lush greenery, with trees and bushes dotting the landscape. The river is bordered by a dirt road, which adds a rustic charm to the scene. The video is taken from a high vantage point, providing a bird's eye view of the river and its surroundings. The overall style of the video is naturalistic, with a focus on the beauty of the landscape and the tranquility of the river."
+    ],
+    "video_ids": [
+        "OOq66pnu08g_3_197to345"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: River, lush greenery, trees, bushes, dirt road? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a river flowing through a landscape with lush greenery, trees, bushes, and a dirt road alongside it. The elements described are all present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a river, through a landscape. lush greenery, trees, bushes, and a dirt road. it. The presence are in prominently present and contribute depicted in the scene.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a3b8482767b1bc4bf2b6f0d129e66fad3ade1019
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86c0d99af37248f140f609356916f3d8a1bffcc15427396ec923329010ed25aa
+size 230399
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..30073e58ee699f90cb81b0fc6b4b3e5527c109ce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment from a soccer match featuring a player from the Real Madrid team. The player, wearing a white jersey with the team's logo and sponsor, is seen pointing towards the camera with his right hand. His expression is focused and intense, suggesting he is communicating with his teammates or strategizing during the game. The background is filled with the blurred figures of other players and spectators, indicating the match is taking place in a stadium. The style of the video is dynamic and action-oriented, capturing the energy and excitement of the sport."
+    ],
+    "video_ids": [
+        "z3WPrzDM-MM_24_196to376"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player from Real Madrid wearing a white jersey with the team's logo and sponsor.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player wearing a white Real Madrid jersey with the team's logo and the sponsor 'Emirates' clearly visible. The player's attire and the context (football match, stadium crowd) align with the description, confirming the core condition is met.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player wearing a white jersey Madrid jersey with the team's logo and sponsor sponsor 'Betirates' visible visible. The player is attire and the context ofa match) crowd)) align with the description of fulfilling that fulfillment condition is met.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7414509a2bd193b21a73f9156241d7f67f1aea12
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e6061583385152ef73e0433dbbdeb48f5cd44dd2e2b0787b8e6a9146f05ef3df
+size 170021
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..2c39b1cc0e281c15ecbf726fdc1b5a2a748e8e05
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video opens with a close-up shot of a hummingbird hovering near vibrant purple flowers. The bird's wings are in rapid motion, creating a blur effect as it feeds on the nectar. The flowers are densely packed, with their delicate petals and slender stems filling the frame. The background is softly blurred, drawing attention to the hummingbird and the flowers. As the camera slowly zooms out, the scene transitions to a wider view, revealing a vast field covered in these same purple flowers. The hummingbird continues its flight across the field, its wings still flitting rapidly. The camera follows the bird, capturing its graceful movements against the backdrop of the expansive floral landscape. The overall atmosphere is serene and natural, highlighting the beauty of the hummingbird and the vivid colors of the flowers."
+    ],
+    "video_ids": [
+        "3d0d50cefa4000cef6274f589e3c58d2d4d0ca13050e5aae2a66d6f4ee8e7017"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Hummingbird, Purple flowers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a hummingbird interacting with purple flowers. The hummingbird is visible in flight and perched among the flowers, and the purple flowers dominate the visual field, matching the description accurately.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as featuring showing a hummingbird and with purple flowers. The hummingbird is depicted in the, isched on the flowers, and the flowers flowers are the scene field, matching the description provided.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d322ecbf3fb5ec79edf549886abc3e9eccbcaf28
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77d9ce4196e30a179b59dda56ee7c9dfe8d26a755d358a94b8084b4785a3cd4b
+size 358830
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..33f3e5b391492ef26cb53efc5d481ea1f3e71a42
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a drone shot of a tropical beach resort. The first frame shows a wide shot of the resort with a clear blue sky and turquoise ocean. The second frame zooms in on the beach, revealing a row of colorful umbrellas and lounge chairs. The third frame shows a closer view of the resort buildings, tennis courts, and a boat docked in the water. The style of the video is aerial and dynamic, showcasing the beauty of the resort and its surroundings."
+    ],
+    "video_ids": [
+        "p9jlvEtctgM_58_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A row of colorful umbrellas and lounge chairs, resort buildings, tennis courts, and a boat docked in the water.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a row of colorful umbrellas and lounge chairs on the beach, resort buildings along the shoreline, tennis courts adjacent to the beach area, and a boat docked in the water. The visual elements align with the described objects without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showcasing showing a row of colorful umbrellas and lounge chairs, a beach, resort buildings, the shoreline, and courts, to the beach,, and a boat docked in the water. The presence elements in with the description objects, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f418787c0a3ea76d4a130d6126c7c91ba078963b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1208-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c37c033a2aa3cc5ea5978f5409d851c55a50dd18d9e6ea0347e71b59a9a6f8d1
+size 281734
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..c291c87ff0a8251a22f0f0970981865ae6c77e91
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in a snowy field, wearing a black coat with a hood. He appears to be speaking or listening intently, as he is looking to the side with his eyes closed. The background shows a white house with a wooden fence and a line of trees. The style of the video is a straightforward, real-life scene with no special effects or artistic filters. The focus is on the man and his interaction with the environment."
+    ],
+    "video_ids": [
+        "TvNdbe-QwwI_5_0to113"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man wearing a black coat with a hood, looking to the side with his eyes closed.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black coat with a hood, and he is looking to the side with his eyes closed for a significant portion of the clip. The background elements (trees, fence, house) do not contradict the core description and are acceptable as additional context.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black coat with a hood, which his appears indeed to the side with his eyes closed. a significant portion of the frames. The background includes,snow and snow, and) do not contradict the description description and are acceptable additional additional elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c22138c7fa599815c5b689dd6aa6e56f3d922458
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:22fbd7100d8ca9eb9f86e5252b6ea790ba0827966ae8bae47b5298c9f0208c3c
+size 148878
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..c113954110f2323eefabf56e7b45b47d6128b0b1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen in a motorcycle gear store, standing in front of a display of helmets and jackets. He is wearing a gray t-shirt with the number 38 on it and is gesturing with his hands, possibly explaining something to the camera. The store is filled with various motorcycle gear, including helmets and jackets, which are displayed on the walls and shelves. The man appears to be the main subject of the video, as he is the only person visible in the frames. The style of the video seems to be a casual, informative tour of the motorcycle gear store, with the man providing commentary on the products."
+    ],
+    "video_ids": [
+        "WrNDej13E1Q_20_44to225"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in motorcycle gear, wearing a gray t-shirt with the number 38, standing in front of a display of helmets and jackets.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a gray t-shirt with the number 38, standing in front of a display that includes helmets and jackets, which matches the core description. Additional elements like a motorcycle and another person in gear do not contradict the main condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a man standing a gray t-shirt with the number 38, standing in front of a display of includes various and jackets. which align the description description. The elements like the microphone helmet a person in the are not contradict the main description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e6ee5798b2316d1876d36557b8542450eeda8141
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:427a004b60842bba264104ac848041bb4d79636ad4c3ad0e42937fcc10c00e6b
+size 222678
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..13b4b39481e5b377d371f3fd8fc07fe556797183
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen holding a pink tutu dress with a silver bow on it. She is standing in front of a window, which is letting in natural light. The woman is wearing a white sweater and a necklace. The video is likely a tutorial or a review of the tutu dress, as the woman is examining the dress closely. The style of the video is casual and informative, with a focus on the tutu dress."
+    ],
+    "video_ids": [
+        "7qZ2gbOm20E_0_26to185"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a pink tutu dress with a silver bow, a white sweater, and a necklace.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman holding a pink tutu dress with a silver bow, wearing a white sweater and a necklace. These elements match the description provided in the 'Object(s)' condition, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a pink tutu dress with a silver bow, which a white sweater, a necklace. The elements match the description provided, the questionObject(s)' condition. indicating there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..31e067c912ad572433de1e8e611af9bb5fb9886d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b068df027d4861233948ba64eebd54fceb99b540266996190e1dd1474a685a9
+size 204056
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..011c36a93648252dc823e7b14c55ba73636c8738
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is an aerial view of a town in Chile, specifically Puerto Varas. The town is situated on the shore of a large body of water, possibly a lake or a bay, with a mountain range visible in the distance. The town is densely populated with buildings of various sizes and colors, including red, yellow, and white. The streets are lined with trees and grassy areas, and there are cars parked along the sides of the roads. The overall style of the video is a realistic, color-rich depiction of a small town in a beautiful natural setting. The video captures the town's layout and the surrounding landscape, providing a comprehensive view of the area."
+    ],
+    "video_ids": [
+        "9E2TjdcLep8_7_0to153"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Buildings, streets, trees, grassy areas, cars? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing buildings, streets, trees, grassy areas, and cars in the aerial view of Puerto Varas, Chile. These elements are prominently featured and accurately represent the described objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing buildings, streets, and, grassy areas, and cars. a scene view of a Nas. Chile. The elements are present featured and do represent the described scene.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b871d23fee69d95f01feff9ac140da61ec566e3b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:31d709168f84832df6cd6a8063d0a2137938aea90cb90b5303caf48d6cb3b578
+size 161486
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..fbc8c2f1a3e8cb6d60c4c3325beb64501ee13034
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a red dinosaur toy with sharp teeth and a fierce expression, standing in front of a fake rock formation. The dinosaur is the main focus of the video, and it appears to be in motion, possibly walking or running. The background consists of a blue sky with white clouds, and the rock formation is made of a light-colored material that resembles stone. The style of the video is playful and cartoonish, with a focus on the dinosaur toy. The lighting is bright and even, suggesting that the video was taken in a well-lit indoor setting. The overall impression is that of a fun and imaginative playtime scenario."
+    ],
+    "video_ids": [
+        "2R6QabVy-Ek_19_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red dinosaur toy with sharp teeth and a fierce expression.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red dinosaur toy with sharp teeth and a fierce expression, matching the description. The toy's mouth is open, revealing detailed teeth, and its facial features convey a fierce look. Additional elements like the background scenery do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a red dinosaur toy with sharp teeth and a fierce expression, which the description provided The toy is design is open, displaying its teeth, and it eyes expression convey a fierce demeanor. The elements like the blue and and not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9bff46ce490b0cc4daf3844bde027b3a54698344
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f4a71f0c3a0baaf314c65fb6c700d2c4582209a37a3f63b49f793d31094a4aa
+size 189861
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..85167c4ef9421d56c2f1e4e357f332f38a6848a2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a white boat being towed on a trailer by a truck. The boat is large and has a canopy. The truck is driving down a street with buildings on both sides. The buildings have blue awnings. There are palm trees in the background. The sky is blue with some clouds. The truck is pulling the boat towards the right side of the screen. The boat is on the trailer and the trailer is attached to the truck. The truck is moving and the boat is being towed. The boat is white and the trailer is black. The truck is pulling the boat and the boat is on the trailer. The trailer is attached to the truck and the truck is pulling the boat. The boat is being towed by the truck. The truck is pulling the boat and the boat is on the trailer. The trailer is attached to the truck and the truck is pulling the boat. The boat is being towed by the truck. The truck is pulling the boat and the boat is on the trailer. The trailer is attached to the truck and the truck is pulling the boat. The boat is being towed by the truck. The truck is pulling the boat and the boat is on the trailer. The trailer is attached to the truck and the truck is pulling the boat. The boat is being towed by the truck. The truck is pulling the boat and the boat is on the trailer. The trailer is attached to the truck and the truck is pulling the boat. The"
+    ],
+    "video_ids": [
+        "O_jCWK9Yw1w_18_0to123"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white boat, a truck, a black trailer, buildings with blue awnings, palm trees, and a blue sky with clouds.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a white boat on a trailer, buildings with blue awnings, palm trees, and a blue sky with clouds. The trailer is silver, not black, but this is a minor deviation from the description and does not contradict the core elements. The presence of a truck is not explicitly shown, but the trailer is attached to a vehicle, which is implied. The video largely matches the description without major conflicts.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a white boat, a black, a with blue awnings, palm trees, and a blue sky with clouds. The truck is black, not black, but the does a minor discrepancy from the description. does not significantly the core elements. The presence of a truck is not explicitly mentioned, but the trailer and being to a vehicle, which could implied to The overall captures matches the description provided any contradictions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f5e6ac7265a67a15ecf776a555338d28b08e631e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f242f2f2871a475011aca299cc66cc1fbf9769db89577e8a29a8dffd5ca4093
+size 303380
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..c5a8e523f984cf36a660b9928d94876fcd25fd4d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated depiction of a vintage fire truck on a track, with two animated firemen in blue uniforms and yellow helmets. The fire truck is red with a gold grill and a large bell on top. The firemen are standing on either side of the truck, one on the left and one on the right. The truck is on a track that is surrounded by grass and rocks. In the background, there is a brick building with a large door and a window. The style of the video is cartoonish and colorful, with a focus on the fire truck and the firemen. The animation is smooth and the colors are vibrant. The video does not contain any text."
+    ],
+    "video_ids": [
+        "cy14sJz-ltg_36_0to145"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A vintage red fire truck, two firemen in blue uniforms and yellow helmets? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a vintage red fire truck with a hose reel on top, consistent with the description. Two firemen in blue uniforms and yellow helmets are visible; one is driving and the other is seated beside him. The scene also includes a brick building and a railway track, which do not contradict the core description and are acceptable as additional elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a vintage red fire truck with a gold attached on the, which with the description. There charactersmen in blue uniforms and yellow helmets are also, one is interacting the the other is standing in the. The setting is includes a building building and a grass track, which are not contradict the core description but are additional additional additional elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bb17a784b347cae0300803b30b9478575255aaf5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0faf9c7e9600bba63d67964a783c7e39a5ef57dfd3cb9c328f167ad4df642c66
+size 208816
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..8ae412fbb5955cf9d29ede8b7fe9bec38f5cad12
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a 3D animated brain with a realistic texture and color, set against a blue background. The brain is shown in three different positions, each time with a slightly different angle and perspective. The animation is smooth and fluid, giving the impression of the brain moving or rotating. The style of the video is educational and informative, likely used to illustrate the structure and function of the brain."
+    ],
+    "video_ids": [
+        "RQVQrT3-vPU_0_0to193"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: 3D animated brain? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video displays a 3D animated brain rotating against a blue background, matching the description of a 3D animated brain. There are no conflicting elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts two 3D animated brain rotating against a blue background. which the description of a 3D animated brain. There are no additional elements in would the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..89027991e3fa2a17f2fc2a9ae4f0bfe4f08a3b09
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1209-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5862eda548b30cee7f6ed853ef7d1d8dc0801c6dbcc36c9e73e2276d5aa8a06e
+size 243087
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..0bf824ab800b07369c38d60eeea77e0a77c26ec4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man's journey along a rocky beach. He is dressed casually in a black t-shirt and sunglasses, carrying a backpack. The man is seen walking towards the camera, his gaze fixed on the distant horizon where a mountain range meets the ocean. The beach is dotted with a few other people, some of whom are enjoying the water. The sky above is clear, suggesting a sunny day. The overall style of the video is a simple, unobtrusive documentary, capturing the man's journey and the natural beauty of the beach and ocean."
+    ],
+    "video_ids": [
+        "ck0HqHk1RH4_18_0to212"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black t-shirt and sunglasses, carrying a backpack; a few other people on the beach.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a man wearing a black t-shirt and sunglasses, with a backpack visible on his shoulders, which matches the core description. Additionally, there are other people on the beach, including individuals near the water and under umbrellas, which aligns with the secondary condition. The scene is consistent with a beach setting, and no elements contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a man wearing a black t-shirt and sunglasses, carrying a backpack, on his back. which align the description description. Additionally, there are a people visible the beach in which a in the water and others arellas, which aligns with the description part. The setting is set with the beach setting, and the elements contradict the given.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..001b1464e641fb40c19efdb08d468893736df922
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:677fde5d9369f36d0cf7502bc5a80ee8d048dfd43870d9e8a75c9fcf4d8e08ae
+size 231349
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..c680647f2fd785cdd918cd61f49dea08d255b2e6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a bowl of soup being prepared. The soup is filled with various ingredients, including meatballs, herbs, and vegetables. The ingredients are being stirred and mixed together, creating a delicious and hearty meal. The soup is served in a white bowl, which contrasts nicely with the colorful ingredients. The overall style of the video is simple and straightforward, focusing on the preparation of the soup rather than any additional elements. The video does not contain any text or narration, allowing the viewer to focus solely on the soup and its preparation."
+    ],
+    "video_ids": [
+        "PvX-ZgdDW20_22_0to152"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Ingredients (meatballs, herbs, vegetables) in the soup; a white bowl serving the soup.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white bowl filled with soup containing visible meatballs, chopped green herbs (likely chives or scallions), and other ingredients that appear to be meat or vegetables. The composition matches the described elements without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close bowl containing with a that ingredients ingredientsballs, herbs herbs herbs,likely parsleyives or parsleyallions), and pieces vegetables that appear to be vegetablesballs vegetables. The presence and the description ' of any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..933080c6b5ab4de624a887238734bdab62977819
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:69656962a3b6fda5b5b3fc2178b80b8aba86bc052e5e44f00c518f042e73e60c
+size 130217
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..60b9df4796e9028635d61e3ef5baf9b5ebf9eaec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man enjoying a meal at a restaurant. He is seated at a table, wearing a black t-shirt, and appears to be in the middle of a conversation. The man is holding a sandwich in his hands, taking a bite, and savoring the taste. The sandwich is generously filled with ingredients, suggesting it might be a sub or a burger. The restaurant has a casual atmosphere, with other patrons visible in the background. The lighting is warm and inviting, creating a comfortable dining environment. The man's enjoyment of his meal and the relaxed setting suggest a pleasant dining experience."
+    ],
+    "video_ids": [
+        "RYHs4tV3vcY_23_62to242"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a table, a sandwich? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man eating a sandwich at a table in what appears to be a casual dining setting. The core objects \u2014 a man, a table, and a sandwich \u2014 are all present and accurately depicted. Additional background elements (other people, ceiling fans, etc.) do not contradict the description and are consistent with a public dining environment.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man sitting a sandwich at a table. a appears to be a restaurant dining setting. The presence elements ( a man, a table, and a sandwich \u2014 are all present and fulfill depicted. The elements elements likelike din, restaurant lights, etc.) do not contradict the main and are typical with a typical dining environment.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ca16ec5d90470c024cd7d73efeb3dffa3a003d57
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7265a8d34f4a484c4ecd72056936c326616013eb827157a2bb0391c19540635d
+size 133503
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..60d9c411d86b8f8b1ad87f443f70a5d741019075
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with short blonde hair and glasses, wearing a black and white scarf. She is seated at a table in a room with a window and a painting on the wall. The woman appears to be deep in thought, with her eyes closed and a serious expression on her face. The room has a warm and inviting atmosphere, with natural light coming in through the window. The painting on the wall adds a touch of elegance to the room. The woman's scarf adds a pop of color to her outfit. The overall style of the video is realistic and naturalistic, capturing a candid moment in the woman's life."
+    ],
+    "video_ids": [
+        "-uXbyqJQMns_21_0to154"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with short blonde hair and glasses, wearing a black and white scarf, sitting at a table. A painting on the wall.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with short blonde hair and glasses, wearing a black and white patterned scarf, sitting at a table. In the background, there is a visible painting on the wall, matching the description. The core elements are accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with short blonde hair and glasses, wearing a black and white scarfed scarf, sitting at a table. There the background, there is a painting painting on the wall, which the description provided The presence elements of present represented in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7fc35aa051a5e575208b8521842fdcfef70d6913
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2a1b3f5be4f060dc8a622a23304f22f54d247e18635d79b400e0241edca9c85d
+size 205626
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..c09fde2f8dbb04cfd80619208c0a59a8bca73db2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a large, light brown rhinoceros standing in an indoor enclosure. The rhino is facing the camera, its head slightly tilted to the side, and its eyes are open. The enclosure has a white ceiling with green beams and a white wall. There is a window in the background, allowing natural light to enter the space. The rhino appears calm and is the main focus of the video. The style of the video is a straightforward, unedited capture of the rhino in its environment, with no additional elements or effects."
+    ],
+    "video_ids": [
+        "8CX-6RONM9M_12_49to185"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large, light brown rhinoceros? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large, light brown rhinoceros with a prominent horn, which matches the description. The rhino is the central focus and its features are consistent with the specified object.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a large, light brown rhinoceros. a smooth horn, large matches the description of The rhin is depicted central focus of there size, clearly with the typical object.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7eba086efbcfeef26ae472046c41de6b531017ca
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8bb6f66159d53f0f2590bba4a00f0309776543b71554e5f9b14f9bcb98c0e454
+size 178335
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..9259d2537040a2700a40dea92f8e8934365e2aef
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a stylized, animated depiction of a group of superheroes. The characters are drawn in a dynamic, comic book style with vibrant colors and detailed costumes. The characters are standing in a line, each with a unique pose and expression, suggesting they are ready for action. The background is a dramatic, shattered wall with debris flying around, indicating a recent battle or conflict. The overall style of the video is energetic and heroic, with a focus on the characters' strength and readiness to face any challenge."
+    ],
+    "video_ids": [
+        "M5xs_BFYouc_4_0to173"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A group of superheroes standing in a line, each with a unique pose and expression.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group of superheroes standing in a line, each with a unique pose and expression. Characters like Superman, Green Lantern, The Flash, and others are depicted with distinct stances and facial expressions, fulfilling the 'Object(s)' condition. Additional elements like lightning effects and a speech bubble do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a group of superheroes standing in a line, each with a unique pose and expression. The are the, Wonder Lantern, and Flash, and others are depicted with distinct stances and facial expressions, which the 'Object(s)' condition.\"\n The elements like the and and a rocky bubble are not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..49f26a8f2135968a8a3f879c9f8513edea56c3af
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aa5129286e41fbadc0fdb9c355bab520614b61a842eff8b968976aa1305e3e83
+size 114046
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..fd9f03ddf5a56ff70e8431bb5612b5a952020647
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a woman lying in bed, transitioning from a state of wakefulness to sleep. She is wearing a pink shirt and appears to be in a relaxed, comfortable position. The room has a calming blue color scheme, and the lighting is soft and warm, suggesting a peaceful and serene environment. The woman's facial expression changes from a slight frown to a more relaxed and contented look as she drifts off to sleep. The overall style of the video is gentle and soothing, with a focus on the woman's peaceful transition into sleep."
+    ],
+    "video_ids": [
+        "0cwZaoxq-ns_117_0to129"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman lying in a bed.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman lying in a bed, appearing to be asleep. The scene is consistent with the description, with no conflicting elements. The woman is positioned on a bed with pillows and bedding, and the setting appears to be a bedroom. All visual elements support the core description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a woman lying down a bed, which to be asleep. She setting is consistent with the description of with the additional elements present The woman is wearing on a bed with a, a, and her setting suggests to be indoors bedroom at The these elements align the ' description.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8970fe8eb8ab5459234c21e04a31081b4dcb5c42
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a67d46f4f7137b4443f7815c729dbf1fc11ab1f738fe80180ac28de29f1f5977
+size 44536
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..514529ab4bed3f1fa6df88589ef9a6df98cbb9b7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young woman with glasses is sitting at a wooden dining table, engaged in a conversation with an older man. The woman is wearing a black t-shirt and has her hair pulled back. She is holding a glass of water in her hand. The older man is wearing a white shirt and is seated across from her. The table is set with a plate of food and a glass of water. The scene appears to be taking place in a home setting, with a window in the background. The video captures a candid moment of conversation between the two individuals."
+    ],
+    "video_ids": [
+        "5iIq0s6iP4c_49_0to122"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman with glasses and a black t-shirt, an older man in a white shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman with glasses and a black t-shirt seated at a table, and an older man in a white shirt is partially visible to her right. These elements match the description provided, and no contradictory elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman wearing glasses and a black t-shirt sitting at a table. holding an older man in a white shirt is partially visible on her right. The elements match the description provided, indicating the additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7c5d15ba701d90be507bea7246d5ff4cb106610b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-121-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd2f530584c4774765ca170f3b067d42deb07629da08d784668d81247b9858b8
+size 120492
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..a5dd092a3b36b3b2ec5bcb12625df89bad43c913
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a slice of chocolate and peanut butter pie being served from a larger pie. The pie is placed on a wooden table, and the slice is placed on a white plate. The pie is topped with chocolate and peanuts, and the slice is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate. The pie is placed on a white plate."
+    ],
+    "video_ids": [
+        "jKcfbyT18j4_0_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A slice of chocolate and peanut butter pie, a larger pie, chocolate, peanuts? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a slice of chocolate and peanut butter pie on a white plate, along with the larger pie in the background. Both are topped with chocolate and peanuts, matching the description. The presence of a knife and additional plates does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a slice of chocolate and peanut butter pie on a plate plate, with with a larger pie in the background. The the topped with chocolate and peanuts, which the description. The presence of a fork and the peanuts does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0615f649f3f2975b136c94745c993cbf3a635183
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e85dd437c80ea1f9be1b161eaeb0d7d8410275dd9a6e1dc02ab14f9d4a2453f2
+size 105859
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..696a68690ef4148a1b6991991287791dd1d6023d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a promotional advertisement for the 2020 Porsche Taycan Turbo S. The car is shown driving on a track, with a clear blue sky in the background. The car is white and has black rims. The track is surrounded by a fence and there are trees in the distance. The car is moving at a high speed, and the camera follows it from a side angle. The video is shot in a realistic style, with high-quality footage and clear audio. The car is the main focus of the video, and it is shown in various angles and speeds. The track and the surrounding environment are also visible, but they are not the main focus of the video. The video is designed to showcase the car's performance and design."
+    ],
+    "video_ids": [
+        "TnSWjxxHQXU_0_200to389"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white 2020 Porsche Taycan Turbo S? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a white 2020 Porsche Taycan Turbo S, as confirmed by the on-screen text overlay. The car is clearly visible throughout the video, and its design, color, and model designation match the description. Additional elements like the train and landscape do not contradict the core object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a white car2020 Porsche Taycan Turbo S driving which indicated by the distinct-screen text and. The car is shown visible and the frames, and the design and including, and model are match the description. The elements such the road and the in not contradict the core description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..efba5bae313436b15e12c5cbed0722f115efa9c3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b932b668fbbc3b344f3b47ae75bfc64d4653d9146de8389f453860a4f482c2c3
+size 230051
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..023da82e8afeb7fb83e4f7bebe88725b22413a77
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are seen taking a selfie in front of a scenic lake. They are both wearing sunglasses and smiling at the camera. The man is wearing a gray jacket and the woman is wearing a black tank top. In the background, there are several boats docked at the lake, and a mountain can be seen in the distance. The sky is cloudy, suggesting that it might be a cool day. The overall style of the video is casual and relaxed, capturing a moment of leisure and enjoyment between the two individuals."
+    ],
+    "video_ids": [
+        "crbccDRL4e8_33_0to153"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman standing together, both wearing sunglasses and smiling. They are the central subjects of the video, and no elements contradict the requirement of featuring a man and a woman.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man and a woman standing together, which wearing sunglasses and smiling, The are positioned main focus of the video, and there additional contradict the description of having a man and a woman.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1306a553292f77619e3ab64fec7e7954d6626f90
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7c65cfd2f2d1f97e79adb2b57dcce4a1c254ff495619c4523313aeab061f9077
+size 177119
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..08e4cc7853872e2847b40ed1d66cbe01810cca8c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene lakeside scene where a controlled fire burns amidst tall, dry grass near the water's edge. The flames are vivid and dynamic, casting an orange glow that contrasts sharply with the surrounding greenery and calm water. In the background, a dense forest of evergreen trees lines the shore, adding depth to the tranquil setting. A dog, possibly a golden retriever or similar breed, is seen walking leisurely on the grassy bank, its fur a warm golden hue that complements the natural tones of the environment. The camera remains stationary throughout the sequence, allowing viewers to take in the peaceful yet striking juxtaposition of the controlled fire against the otherwise undisturbed landscape. The overall atmosphere is one of quiet beauty, with the gentle movement of the flames and the dog's slow pace providing subtle motion within the stillness of the scene."
+    ],
+    "video_ids": [
+        "8ec3c2a2e6e8014948865e97035fcf05a8e88658c4f30f110294d12c85e549ee"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A controlled fire burning vividly, a dog walking leisurely on the grassy bank.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a vividly burning controlled fire near a body of water, and a dog is seen walking leisurely on the grassy bank. These elements match the described conditions without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a controlledly burning fire fire in a body of water, with a dog walking seen walking leisurely on the grassy bank. The elements match the description ', any.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..86aa4ac257e5edca7661f15208252dd7a5fa11a5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:09d783b4c38f3c63330541430ea6069be7b6b5a3c80091c0d7f5151b2808d808
+size 233214
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..61084256531e1d6ba93499207b3ab3968cd86d75
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with dark hair and a black top, standing in front of a red building with a white fence. She is wearing earrings and appears to be speaking. The lighting suggests it is daytime. The style of the video is a close-up shot of the woman, focusing on her facial expressions and the building in the background."
+    ],
+    "video_ids": [
+        "Cd90AKobyZY_100_18to201"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with dark hair and a black top, wearing earrings, standing and speaking.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with dark hair, wearing a black top and earrings, standing and speaking. The background elements (e.g., building, shopping carts) do not contradict the core description and are consistent with an outdoor setting.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with dark hair and wearing a black top, earrings. standing and speaking. The background includes,a.g., the and fence cart) do not contradict the core description and are acceptable with the outdoor setting where}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b1fd683036508642603818077d75e3930ddd0020
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8737565e2b2c7de88632012051399c86f6a416f7e63960d6ec06cd23271cafa1
+size 119517
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..50e7dc33858976184665e63f2412357062a9a27d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, three individuals are seated in a casual setting, each wearing a blue shirt with a name tag. They are positioned around a table with various items on it, including a bottle, a cup, and a bowl. The room has a vibrant blue background, and the overall atmosphere is relaxed and friendly. The individuals appear to be engaged in a conversation, suggesting a social gathering or meeting. The style of the video is informal and candid, capturing a moment of interaction among the three people."
+    ],
+    "video_ids": [
+        "1RRumSlNiPg_39_24to211"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three individuals, a bottle, a cup, and a bowl.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows three individuals seated in a studio setting. On the table in front of them, there is a bottle (water bottle), a cup (white disposable cup), and a bowl (white paper bowl with food). These objects are visible and consistent with the 'Object(s)' condition, even though other elements like a phone and magazine are also present, which does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows three individuals, at a room setting. There the table in front of them, there is a bottle,likely),), a cup,with mug cup), and a bowl (white bowl bowl). a remnants The objects match present and match with the 'Object(s)' condition described fulfilling though the elements like the microphone and a are also present on they do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7318ea075397a5631dbd95f70b70f4b21f88aa5e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:291a011f6aa5baabee007c0eea1322b324bffa677c035492a93bfb1511eae9d3
+size 133782
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..6420da81fe2dd30c55712df8af9a70397f6de672
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with glasses, wearing a blue shirt, and a microphone around his neck. He is standing in front of a building with a sign that reads \"IMMIGRANT\". The man appears to be speaking or preparing to speak, as suggested by the microphone. The setting is outdoors, with a fence and a building in the background. The man's expression is serious, indicating that the content of the video might be of a formal or professional nature. The overall style of the video suggests a news report or an interview, with the man possibly being a reporter or an expert on a particular topic."
+    ],
+    "video_ids": [
+        "GONlbGeUmFo_9_0to168"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with glasses, wearing a blue shirt, and a microphone around his neck.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing glasses and a blue shirt with a microphone clipped to his shirt, matching the description. The background elements, such as the building and umbrellas, do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a man wearing glasses and a blue shirt, a microphone in to his shirt. which the description provided The microphone,, such as the building and therellas, are not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9ac535056941838e0c23705ca4dc19254af52a35
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c580bb5ad2f453fa6fe15f409b030efedd7e08373a7dd20ea3f21ba2cf9f939a
+size 209326
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..2d57234e502bcc98e502813ec10c8d4a9f9e6fc6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of triumph and camaraderie among a group of soccer players. The players, all dressed in vibrant yellow jerseys, are seen celebrating a goal on the field. The first frame shows a player in the foreground, his arms raised in a victorious gesture, while his teammates in the background are seen running towards him, their faces lit up with joy. The second frame captures the same moment from a different angle, with the celebrating player now in the background and his teammates in the foreground, their bodies in motion as they join the celebration. The third frame is a close-up of the same player, his face beaming with pride and his teammates' hands on his shoulders, a symbol of their shared success. The video is a dynamic and colorful depiction of teamwork and sportsmanship."
+    ],
+    "video_ids": [
+        "TYmChAG9pfI_27_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Soccer players, a player in the foreground, teammates in the background, and another close-up of the same player with teammates.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple Brazilian soccer players in yellow jerseys celebrating on the field. The camera pans across them, highlighting a player in the foreground (number 16) and then revealing teammates in the background. The sequence includes a close-up of the same player (number 16) with teammates around him, fulfilling the described conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a frames soccer players in yellow jerseys celebrating on a field. The sequence captures to the, showing one player in the foreground withlikely 13) and teammates showing teammates in the background. The sequence also a close-up of the same player withnumber 16) with teammates, him, which the ' conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..eb65636aa26c4a25ebbb748d6a242ee3389a16df
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1210-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6afebcfd15e2df8a8910bb62e56754f181fc0f988315ba96e048fe888ecac83
+size 342794
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..bbaf805821512046469bb8b34486e0d768d05cd2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are seen interacting with a large poster. The man, dressed in a gray shirt, is standing to the right of the woman, who is holding the poster. The poster features a bicycle and a man, and the woman appears to be showing it to the man. The setting is a convention center, with a blue curtain and a white tent visible in the background. The man and woman are the main subjects of the video, and their actions suggest a casual, friendly interaction. The poster, with its bicycle and man, adds a touch of intrigue to the scene. The convention center setting, with its blue curtain and white tent, provides a backdrop that suggests a public event or gathering."
+    ],
+    "video_ids": [
+        "OjxN4Op2bUA_15_31to239"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman, a large poster featuring a bicycle and a man.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a woman interacting with a large poster that features a bicycle and a man, which matches the described condition. The poster is prominently displayed and appears to be the focus of their attention, fulfilling the core requirement.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a woman standing with a large poster. features a man and a man. which align the description '. The poster is prominently displayed, the to be the focus of their interaction, fulfilling the ' description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..72478eb02c386b676bb133d8c21d2ef8603fc4d8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e3741031fe2949dea97008e3c8edce2d2a6d74f028b5de7b93b35a34ecf3227
+size 186618
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..86fae958c0029fd05a571136cda401e4ec40fc03
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a red BMW X2 car on display at a car show. The car is positioned on a black platform, and its shiny exterior reflects the bright lights of the showroom. The car's front grille, headlights, and hood are prominently featured in the video. The car's sleek design and vibrant color make it stand out in the showroom. The video captures the car's details and the overall atmosphere of the car show."
+    ],
+    "video_ids": [
+        "-nmWXStcNBA_71_0to151"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red BMW X2 car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red BMW X2 car prominently displayed, with its distinctive front grille, headlights, and emblem clearly visible. The car's color and model designation are consistent with the description, and no elements contradict this core identification.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red BMW X2 car, displayed, which a front features grille, headlights, and overall clearly visible. The car is design and model are match consistent with the description of and there additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e1dc182f453822e6d0299f24e386863980cb1e3a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a66a30f702ec5340a9fd3056958c66f1cf531140a1b1925ea805fff8e5002f1b
+size 190025
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..5fbfac9a74fb5cd5d68673657333b46493ae42ea
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of triumph for a soccer team, as they celebrate a victory on the field. The team, dressed in vibrant red jerseys, is seen in a group hug, their faces lit up with joy and their arms raised in triumph. The stadium lights shine brightly in the background, illuminating the scene and casting a warm glow on the players. The atmosphere is one of camaraderie and shared success, as the team celebrates their hard-earned victory. The video is a dynamic snapshot of the exhilarating moment of victory in the world of soccer."
+    ],
+    "video_ids": [
+        "s0zn5adUMZY_14_0to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A soccer team in vibrant red jerseys hugging together.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group of soccer players in vibrant red jerseys hugging and celebrating together, which matches the core description. The players are clearly identifiable as a soccer team, and their actions align with the 'hugging together' condition. Additional elements like stadium lights and the 'SPORTUBE' logo do not contradict the main description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three group of soccer players wearing vibrant red jerseys hugging together celebrating on on which align the description description of The players are on wearing by a team team, and their red of with the 'hugging together' part. The elements such the lights and a presenceBS' banner in not contradict the main description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..33fba6d10fb72daf9e27a2eacee042fefdf98b52
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e10482244d36b71297381ea4c645a7b0c5b4b9edfacbdba7ba85e6f5324c9aa
+size 202264
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..616bf0d58d8a6bfc9fd7f7d8123a05d4041c4292
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a red sports car with a yellow license plate, parked in a garage with a green wall and orange flowers. The car is positioned in the center of the frame, and the garage appears to be well-lit. In the foreground, there is a microphone on a stand, suggesting that a podcast or interview might be taking place. The style of the video is realistic, with a focus on the car and the microphone, indicating that the video might be related to a car review or a podcast episode about the car. The overall atmosphere of the video is calm and professional."
+    ],
+    "video_ids": [
+        "8zwVnDcUngA_11_0to104"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red sports car with a yellow license plate, microphone on a stand? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red sports car with a yellow license plate (2 JWW) and a microphone on a stand positioned in the foreground. These core elements are accurately represented, fulfilling the 'Object(s)' condition. Additional elements like the green wall with flowers and speakers are present but do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a red sports car with a yellow license plate,the11  and a microphone on a stand. in front foreground. The elements elements match present depicted, fulfilling the 'Object(s)' condition as The elements like the flowers wall and flowers and the are present but do not contradict the main.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b771bb370cb9fb7055716ef3cf93f76b6615970e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9510ea01005aac5bd24249bd1fa495e67bf9f881d9cbda96896080d424499783
+size 62527
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..a8215119fa2b4c98512e67849658e072f6695062
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a vibrant underwater scene featuring a school of yellow and black striped fish swimming around a coral reef. The fish are darting and darting around the coral, their bodies shimmering in the light filtering through the water. The coral, a mix of orange and red, provides a stark contrast to the blue of the water. The fish are seen swimming in different directions, some closer to the camera, others further away, creating a sense of depth and movement. The video is a beautiful representation of marine life, showcasing the diversity and beauty of the underwater world."
+    ],
+    "video_ids": [
+        "G56T3AMvbG8_166_0to150"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A school of yellow and black striped fish? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a school of fish that are predominantly yellow with black stripes, matching the description. The fish are swimming around a coral structure, and while there are other elements like coral and possibly other fish, the main focus is on the school of yellow and black striped fish, which is accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a school of yellow with are predominantly yellow with black stripes, which the description of The fish are swimming in a coral reef, which the there are other fish like the and additional other types species the main focus is on the yellow of yellow and black striped fish, which ful the represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..30f8682c4e99a8c840cfdf996a90806494f00849
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:356d0d77c8ad60ff07236884a5d4b8510c1e44968de0c019fe4223f14d737d32
+size 317344
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..1f19860e0c252493985bc97d1243c9ffbfa6031c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a fire truck's front bumper, showcasing its various components. The bumper is silver and features a red hose reel with a hose coiled neatly on it. There are also several silver nozzles attached to the bumper, each with a red cap. The fire truck's grill is visible in the background, and it has a shiny chrome finish. The video is a straightforward, unembellished depiction of the fire truck's equipment, focusing on the details of the bumper and its attachments. The style of the video is straightforward and informative, providing a clear view of the fire truck's front bumper and its components."
+    ],
+    "video_ids": [
+        "M6qK43eDrl4_15_0to130"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Fire truck's front bumper, red hose reel, coiled hose, silver nozzles with red caps, fire truck's grill? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows the fire truck's front bumper, red hose reel, coiled hose, silver nozzles with red caps, and fire truck's grill as described. The camera focuses on these elements, and no contradictory elements are present that would invalidate the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful the fire truck's front bumper, a hose reel, coiled hose, silver nozzles with red caps, and the truck's grill. described. The elements angle on these elements, and there additional elements are present.\"\n would alter the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4066a51b9851baa6acfcd3cb2caebeb19ea150ec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:73b8f1bb8f67be19966f1e4784814ac0cd200d8282bc714725fb3df2d71b5d9c
+size 114856
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..81ba7906da2e66036828827e1926e3f1c0db8382
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a motorhome, focusing on the kitchen area. The kitchen features a gas stove with four burners, a sink, and a countertop. The motorhome has a wooden finish, and the kitchen is equipped with modern appliances. The video is likely a promotional or informational video showcasing the features of the motorhome's kitchen. The style of the video is straightforward and informative, with a focus on the practical aspects of the kitchen design."
+    ],
+    "video_ids": [
+        "gkz2Ik_fZhs_3_104to256"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Gas stove with four burners, sink, countertop, modern appliances? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a gas stove with four burners, a sink with a faucet, a countertop, and modern appliances including an oven beneath the stove. These elements are prominently featured and match the description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a gas stove with four burners, a sink, a faucet, and countertop, and modern appliances such a oven and the stove and The elements align consistent featured and align the description provided any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0252da77bee6c9a16de62a330afa72fd71364104
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52e5b2de2f06e6d7b870e3a94ac7d59e361c3bbf254714d166f29f30b7ba06f0
+size 128033
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..92e35e01309aaebdeb12eb56f7411ab4fbb075ff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a group of three friends are gathered in a cozy living room, engaged in a lively conversation. The room is filled with warm, inviting furniture, including a comfortable couch and a sturdy coffee table. The walls are adorned with a variety of posters, adding a personal touch to the space. The friends are casually dressed, reflecting the relaxed atmosphere of the gathering. The video captures the essence of their camaraderie and the joy of spending time together in a comfortable setting."
+    ],
+    "video_ids": [
+        "bpfqYgCvutY_33_0to218"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three friends? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows four individuals in a room, three of whom appear to be friends interacting around a table. While there are four people, the presence of three friends is consistent with the description, as the fourth person may be part of the group or a minor element. The core description of 'Three friends' is largely fulfilled.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three individuals sitting a living, but of whom are to be engaging engaged with a coffee. The there are four people, the core of an individuals is still with the description. as the fourth person could be a of the group or a separate detail in The interaction condition of 'Three friends' is largely fulfilled,}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9ef473867dbd9e3567726fc17ec20ba1589a3998
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1211-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:febfde5cf8eede5c599efd2484a60915a91511dae0a57a113bf68a427e16abeb
+size 234125
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..1bcde055c10e17e9442d901bc23a8cb8fa500a3f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a soccer match. The main focus is a soccer player in a yellow jersey, who is in the midst of a powerful run across the field. He is in motion, with his arms outstretched, embodying the energy and excitement of the game. The player is heading towards the goal, his eyes focused on the target.  The soccer field is a vibrant green, contrasting with the player's yellow jersey. The goal is visible in the distance, adding a sense of depth to the scene. The player's position on the field suggests he is in the attacking phase of the game, ready to strike.  In the background, other players can be seen, each in their own positions, adding to the overall atmosphere of the match. The crowd in the stands is a blur of colors, indicating a large audience, and the stadium lights illuminate the scene, highlighting the action on the field.  The style of the video is dynamic and energetic, capturing the intensity of the game. The focus on the player in the yellow jersey suggests he is a key player in this match, and his actions are likely to have a significant impact on the outcome. The video does not contain any text or additional elements, keeping the focus solely on the action on the field."
+    ],
+    "video_ids": [
+        "7BfKkAY4CUE_25_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A soccer player in a yellow jersey, other players, the goal, the crowd in the stands? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player in a yellow jersey (number 10) celebrating on the field, with other players visible in the background, a goal net in the distance, and a large crowd in the stands. All elements described in the condition are present and consistent with the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a soccer player in a yellow jersey,number 7)) running a the field. with other players in in the background, the goal in, the foreground, and a crowd crowd in the stands. The elements described in the ' are present and match with the scene depicted}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a671df3545f9b6bac6532747bdf267696b83608b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:70d4bb459b50db24b9c021afc3ce4418a0ab6d3ebb480e6b6c7151fb178581df
+size 263352
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..5931e9744d94732bc5060612bf9cd89fd730f347
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man driving a car with a child in the back seat. The man is wearing a blue shirt and glasses. The car is moving down a road with trees on the side. The child in the back seat is looking out the window. The man is focused on the road ahead. The car is a modern model with a sunroof. The road is clear and the weather is sunny. The man is driving carefully and the child is enjoying the ride."
+    ],
+    "video_ids": [
+        "AWBrBy0CuaY_11_0to133"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a child? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in the driver's seat and a child in the back seat, fulfilling the 'Object(s)' condition of a man and a child. Additional elements, such as people outside the car or background scenery, do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man and the driver's seat and a child in the passenger seat of both the 'Object(s)' condition. the man and a child in The elements such such as the walking the car, the scenery, do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..28d715682fb1eff937d859919869359764291d3b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b5502ea69bc58d7152c12227595f10c3c79322477c9b30188214e241666c49d
+size 235141
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..3d59fd55bca1db7930db993c3bfbb6636e97052f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two bearded men are seen outdoors, standing on a wooden deck. They are engaged in a process involving a large metal pot and a funnel. The man on the left is pouring a substance from the funnel into the pot, while the man on the right watches and possibly assists. The deck is surrounded by a natural setting with trees and a hill in the background. The men are dressed casually, and the overall atmosphere suggests a relaxed, outdoor activity. The style of the video is candid and informal, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "1-YrGIy_WXM_22_305to467"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two bearded men, a large metal pot, and a funnel.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two bearded men standing at a wooden deck, one pouring grains from a white bucket into a large metal pot, while the other uses a long stick to stir. The large metal pot is clearly visible and central to the activity. Although a funnel is not explicitly shown, the action of pouring from a bucket into the pot suggests the use of a funnel or similar tool to guide the flow, which aligns with the description. The core elements are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows two bearded men standing outdoors a wooden table. with of a from a metal container into a large metal pot placed which the other man a funnel-hand to stir the The presence metal pot is clearly visible, there to the activity. There there funnel is not explicitly shown, the action of pouring grains a bucket into a pot suggests the use of a funnel, similar device, control the grains, which aligns with the ' of The presence elements of present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..82c5174d6731e78c33b75944bf5c9d9b480296ce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:207d612b02df2cd849e822d33a36e28d0b74a8cd422d85e11a19a3ca739e12f8
+size 189611
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..7c881057a1c5b9f409e5647e4e6a08e3b0815296
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of two glass jars filled with pickled vegetables, placed on a wooden surface. The jars are transparent, allowing the viewer to see the contents inside. The vegetables include slices of red onions and green herbs, which are submerged in a clear brine. The jars are sealed with metal lids, and the contents are arranged neatly within the jars. The style of the video is simple and straightforward, focusing on the pickled vegetables without any additional context or background elements. The lighting is bright, highlighting the colors of the vegetables and the clarity of the brine. The video does not contain any text or narration."
+    ],
+    "video_ids": [
+        "tMSDoDkZN_A_69_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two glass jars filled with pickled vegetables.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two glass jars filled with what appears to be pickled vegetables, including sliced radishes and herbs. Although a hand is seen adding arugula, the core description of two glass jars filled with pickled vegetables is still fulfilled.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two glass jars filled with pick appears to be pickled vegetables, specifically slices onionsishes and possibly. The the specific is visible in moreugula to the core description of the glass jars filled with pickled vegetables is largely fulfilled.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2601fdbc7b7e51d59f561a93466c6b8aa411527b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1553327b98a5af976eee55bdc7af2f26f113033b217d279bc259821a71e01788
+size 140473
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..865b6f99c52bcae66533ef6819b1402b44227fc8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person cutting and eating a piece of ham with a fork and knife. The ham is pink and appears to be cooked, and it is served on a white plate with a red and white checkered tablecloth. The person is using a silver fork and knife to cut and eat the ham. The style of the video is a close-up shot of the person eating, focusing on the food and the utensils being used. The video captures the action of the person cutting and eating the ham, and the textures of the food and the utensils."
+    ],
+    "video_ids": [
+        "4igpC0tAIi4_38_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person, ham, fork, knife, white plate, red and white checkered tablecloth? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person's hands using a fork and knife to cut and lift a piece of ham from a white plate. The plate is on a red and white checkered tablecloth, matching the described elements. The focus is on the meal and utensils, with no conflicting elements present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close using hands using a fork and knife to cut a lift a slice of ham from a white plate. The ham is placed a red and white checkered tablecloth, which the description elements. There presence is on the action preparation theils, with no additional elements present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..26ced00f25ce77ea11762c8e539b5415a6a988af
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68df55dd98504a9bbd47a876d29d3c484d4e68f5c4a1ebd2762a37e1928e62fa
+size 147945
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..0dc3246ed804fe33d14427dfd4e97f9f64f1a769
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the growth of a blackberry bush over time. The first frame shows a small, leafy bush with a few green leaves and a few small, green berries. The second frame shows the bush with more leaves and larger berries, indicating that the bush is growing and the berries are ripening. The third frame shows the bush with even more leaves and larger, darker berries, suggesting that the bush is thriving and the berries are fully ripe. The style of the video is a time-lapse, which allows viewers to see the progression of the bush and berries over time. The focus is on the bush and berries, with no other objects or people in the frame. The video does not contain any text or sound."
+    ],
+    "video_ids": [
+        "QVY4SJt4mzg_69_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Blackberry bush, berries? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bush with green leaves and clusters of small, dark berries, which are consistent with blackberries. The focus remains on the plant and its fruit, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black with green leaves and clusters of berries, dark berries that which align consistent with theberry. The progression is on the berries and berries berries, fulfilling the 'Object(s)' condition of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9de2ccf203edbba11744b4dcd72f60a7c4c10972
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3268d8e17f846266a03dfd1b96dff928bb867370e9b7e88e4657ac06ba568533
+size 224855
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..9eeb3875fcd81dffffcb19855249e1d8742b9647
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up view of the interior of a luxury car, focusing on the steering wheel and dashboard. The steering wheel is black with a silver emblem in the center, and it has multiple buttons on the spokes. The dashboard is sleek and modern, with a digital display showing the car's speed and other information. The car's interior is well-lit, with the sun shining through the windshield, casting a warm glow on the dashboard and steering wheel. The style of the video is a professional car review, with a focus on the car's interior features and design. The video is likely intended to showcase the car's luxury and high-end features to potential buyers."
+    ],
+    "video_ids": [
+        "Cg4wvCl7HQk_19_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, dashboard? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features the steering wheel and dashboard of a car, specifically highlighting the Abarth logo on the steering wheel, the instrument cluster, and the infotainment screen. These elements clearly fulfill the 'Object(s)' condition as described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a steering wheel and dashboard of a car, which highlighting the Mercedes-Classth logo on the steering wheel and which digital cluster, and the overallotainment system. The elements clearly fulfill the 'Object(s)' condition as they.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e5cb145677e3223c848aa0268b4b4d287cd3ee45
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f2d11e92421c3044fa6149f89954088e05b64cb0c9fa394328a202412a3b4b3f
+size 148187
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..0e66d10f701185378d6b666d273481bddc3cdd88
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young girl with blonde hair is seen sitting at a dining table in a room with a window. She is wearing a pink shirt with ruffles at the sleeves. In the first frame, she is looking at a pink tablet with a pink case on the table in front of her. In the second frame, she is still looking at the tablet, but her expression has changed to one of surprise or shock. In the third frame, she is still looking at the tablet, but her expression has returned to one of curiosity or interest. The room has a cozy and comfortable atmosphere, with a lamp and a picture on the wall in the background. The video captures a moment of discovery and learning for the young girl."
+    ],
+    "video_ids": [
+        "DlsOXcWWkFY_113_0to130"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young girl with blonde hair, a pink tablet with a pink case.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young girl with blonde hair sitting at a table, holding a pink tablet with a pink case. The core elements described are present and accurately depicted, with no contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young girl with blonde hair wearing at a table, interacting a pink tablet with a pink case. The setting elements of in present: match depicted in with no additional or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4a0b69a0dc493936dc1beba7a8c0289460655bdd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1212-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec9ba9e35f94187b8b1ff21750da6a546f9982c9440a816298c3bff42acaf9a6
+size 88954
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..e5ae712ca0d2f2577cb75a5bae569629b90ff844
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a nighttime scene of a McDonald's restaurant. The restaurant is illuminated with bright yellow and red lights, and the iconic golden arches are prominently displayed. The interior of the restaurant is visible through the large windows, showing a clean and well-lit space. The exterior of the restaurant is adorned with a red and white striped awning, adding to the vibrant color scheme. The video is likely a promotional or informational piece, showcasing the restaurant's exterior and interior lighting. The style of the video is straightforward and informative, with a focus on the restaurant's branding and ambiance."
+    ],
+    "video_ids": [
+        "09R151e5Ukw_14_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: McDonald's restaurant with golden arches, large windows, interior space, and red and white striped awning.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a McDonald's restaurant at night, featuring the iconic golden arches, large windows that reveal the interior space, and a red and white striped awning. These elements align with the description provided, even though additional graphical overlays (like the number '33' and logo) are present, which do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a McDonald's restaurant with night with featuring the iconic golden arches, large windows, provide the interior space, and a red and white striped awning. The elements align with the description provided, making though the elements elements orlike the ' '11' and ' ' are present, they do not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2644015f5adcb7f2b6467cf59d1757c69f73ff58
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dfd9a506644ff01a2f26836e1c7c3ed9418b8becad613fc9b7a0e5794d30a58e
+size 152356
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..3850e8bd177b61200e4e27feb25bfc15dd543dd1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic scene on a soccer field. A group of players, dressed in blue and red jerseys, are engaged in a game. The players are in motion, running and chasing after the soccer ball, which is in the air. The field is lush and green, providing a stark contrast to the players' colorful jerseys. In the background, a man in a red shirt is watching the game intently, possibly a coach or a referee. The overall style of the video is energetic and action-packed, capturing the intensity and excitement of the game."
+    ],
+    "video_ids": [
+        "FRxDI0gQRWQ_9_37to218"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Players in blue and red jerseys, a soccer ball, a man in a red shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows players in blue and red jerseys (some wearing orange training vests over blue jerseys), a man in a red shirt (likely a coach), and although a soccer ball is not clearly visible in the frames, the context of a soccer training session implies its presence. The core elements described are present and consistent with the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts players in blue and red jerseys,though with blue shirts bib), their jerseys), a soccer in a red shirt (likely a referee or and a the soccer ball is not clearly visible in the provided provided the context suggests the soccer game session is its presence. The video elements of in present, the with the video depicted}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a89cf1e67c18816f15594f42140a120211d7b563
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ca284ab4d9a6fa178f7b3ae1a7bae2b061b0ae05fa904419e426b34fc3e17334
+size 228361
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..30150cb96d069a42797bd819fd2fc6e15e275884
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a basketball player in action, wearing a red jersey with the number 25 and the word \"Sixers\" on it. The player is seen in three different frames, each showing him in a different position on the court. In the first frame, he is seen running towards the basket, his eyes focused on the ball. In the second frame, he is seen jumping up to catch the ball, his arms outstretched. In the third frame, he is seen shooting the ball towards the basket, his body leaning forward in anticipation. The background of the video shows a blurred image of a basketball court, indicating that the player is in the middle of a game. The style of the video is dynamic and action-packed, capturing the intensity and excitement of a basketball game."
+    ],
+    "video_ids": [
+        "iaQ3Ec8KUS4_13_0to102"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A basketball player in a red jersey with the number 25 and 'Sixers' on it.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a basketball player wearing a red jersey with the number 25 and the word 'Sixers' clearly visible on the front, matching the description. The player is holding a basketball and appears to be in a game setting, which aligns with the 'Object(s)' condition. There are no conflicting elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a basketball player wearing a red jersey with the number 25 and the word 'Sixers' on visible on the front. which the description provided The player is also a basketball and appears to be on a basketball setting, which aligns with the contextbasket(s)' condition.\"\n The are no elements elements in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4da07cde631e4b94014fc16c01cff095957cb056
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:879459de71bbdb22f961323ad7b4ca6738d2148d3e9770d615c78114776aab50
+size 255264
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..bf0ff6c8e57ce14bb98226961211a7a29ef5a139
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red sports car with a black carbon fiber rear wing. The car is parked on a grassy area with a cloudy sky in the background. The car's design is sleek and modern, with gold rims on the wheels. The car's body is shiny and reflects the surrounding environment. The car's rear wing is a prominent feature, adding to its sporty appearance. The car is stationary, and there are no people or other objects in the immediate vicinity. The overall style of the video is a close-up shot of the car, focusing on its design and details."
+    ],
+    "video_ids": [
+        "UrmPokD8Fug_25_0to152"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red sports car with a black carbon fiber rear wing, gold rims on the wheels? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red sports car with a black carbon fiber rear wing and gold rims on the wheels, matching the description. The camera slowly zooms in on the rear of the car, highlighting these features without any conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a red sports car with a black carbon fiber rear wing and gold rims on the wheels, which the description provided The car angle zooms in on the rear of the car, highlighting these features. any additional elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..55f2f8bc73c432dfb0d92bc73cf583245e5a92e7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b226d5cb175b2bd47cd64cba35d0ff49e673424e3c153918dc07c71d47ccff4
+size 289475
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..07472dae12f3b788e249ae84e8a2a8ffb331dd4e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young woman with long blonde hair is seen wearing a fur-lined hooded jacket, smiling and engaging in conversation with an older woman. The older woman, wearing a red jacket with a floral scarf, appears to be listening attentively. The setting is outdoors, with trees in the background, suggesting a park or a similar natural environment. The young woman is holding a microphone with an ABC logo, indicating that she might be conducting an interview or a news report. The overall style of the video suggests a casual, friendly interaction between the two women, with the focus on their conversation."
+    ],
+    "video_ids": [
+        "T3w6WFJJ2wQ_19_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman with long blonde hair, an older woman wearing a red jacket with a floral scarf, and a microphone with an ABC logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman with long blonde hair wearing a fur-lined hood and a maroon jacket, standing next to an older woman in a red jacket with a floral scarf. A microphone with the ABC logo is visible in front of the young woman, matching the description. The background and other elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two young woman with long blonde hair and a p-lined jacketed a microphoneoon jacket, holding next to an older woman wearing a red jacket with a floral scarf. The microphone with an ABC logo is being in the of the young woman, indicating the description provided The setting includes the elements in not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e7ef4a8c2662de79ddfa7a6b4f5585d1cbc8661c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79d9e6382b043415abbef188a4572522107be29a0143252824f303e8fdf44e1f
+size 189824
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..97175f11e6dae72aceba538f8fe55220d6e8300f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a journey on a gravel road through a mountainous landscape. The road is flanked by a river on one side and a mountain range on the other. The vehicles, including a truck and a bus, are seen driving along the road, with the river flowing parallel to them. The mountains in the background are majestic and covered in snow, adding to the scenic beauty of the journey. The overall style of the video is a travelogue, showcasing the natural beauty of the region and the journey itself."
+    ],
+    "video_ids": [
+        "VyKgSnlOayM_56_0to166"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A truck and a bus are driving on the road.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two vehicles, one orange and one white, driving on a dirt road. While they are not explicitly labeled as a truck and a bus, their shapes and sizes are consistent with those types of vehicles. The core description of vehicles driving on the road is accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a vehicles, a that and one black, driving on a road road. The the are not explicitly described as a truck and a bus, the size and sizes suggest consistent with these of of vehicles. The presence condition of the driving on a road is fulfilled represented in}<|im_end|>\n addCriterion",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ca64390d8ac26e5edc8055e7b3ccae39c4597456
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:635af45f9573b71a92465e02ba9efe54846ee872740bf0dac4500fab548bfea1
+size 218522
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..d7f5169a98c988391d4caa26e4d2d4805eabe76f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a thrilling moment in a football game. The scene is set in a stadium filled with spectators, their faces a blur of anticipation and excitement. The main focus is on two players, one from the Cowboys and the other from the Redskins, both in the midst of an intense play.  In the first frame, the Cowboys player, donned in a white jersey, is seen running towards the end zone, his eyes focused on the goal. His body language suggests a sense of urgency and determination.  The second frame captures the moment when the Redskins player, wearing a maroon jersey, leaps into the air, his arms outstretched to intercept the ball. His athletic form is a testament to the agility and speed required in the sport.  The third frame shows the aftermath of the play. The Redskins player, having successfully intercepted the ball, is seen holding it securely in his hands. The Cowboys player, on the other hand, is seen looking up at the sky, perhaps in disbelief or frustration at the turn of events.  The video is a dynamic snapshot of a football game, capturing the high-stakes action and the split-second decisions that define the sport."
+    ],
+    "video_ids": [
+        "1oVL5g32TKw_52_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two football players - one in a white jersey (Cowboys) and one in a maroon jersey (Redskins).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two football players: one in a white jersey with blue accents (Dallas Cowboys) and another in a maroon jersey with yellow accents (Washington Redskins). Both players are clearly visible and identifiable by their team colors and uniforms, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two football players, one in a white jersey and the and,Cow Cowboys) and one in a maroon jersey with white accents (Washington Redskins). The players are engaged visible and are by their uniforms colors and uniforms. fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9d8e7039c18ad404bc614694031d7d1c1ecf52bd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:437cf28834abe767f43984fb8b5e4d0f3b2bde169d91777f635a261d3838a584
+size 351638
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..eeb8e30cea4a58e1fb5c43f29a761c10fe602fa0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen sitting on a bed in a cozy room with a wooden wall. She is wearing a pink shirt and is engrossed in writing in a notebook. The room is filled with warm light, creating a comfortable and inviting atmosphere. The woman's focus on her writing suggests a sense of tranquility and introspection. The wooden wall and the bed add to the rustic charm of the room, making it a perfect setting for quiet reflection and creativity."
+    ],
+    "video_ids": [
+        "4Tg2ZibXb4w_16_17to166"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman sitting on a bed, wearing a pink shirt, writing in a notebook.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman sitting on a bed, wearing a pink shirt, and actively writing in a notebook. The scene is consistent with the described condition, with no significant contradictions. Additional elements like the wooden wall and photos in the background do not conflict with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman sitting on a bed, wearing a pink shirt, and writing writing in a notebook. The setting is consistent with the description scenario, with no additional additional or The elements like the wooden wall in the in the background do not detr with the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5633f0ea5875431d8255e8f4a45ae5f30d3f9d62
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1213-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a86b45e1f7b013ce9de3aeb2f83b5e7200fba3e52ae8fd0a280a26eebdc4405c
+size 78925
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..1b3623329aca9533e9dcbbe1ae0fd91d76df080d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young girl with dark hair, who is the main subject. She is wearing a colorful, striped tank top. The girl is seen in three different frames, each capturing her in a different pose or expression. The background is a plain, light-colored wall, which provides a neutral backdrop that highlights the girl. The style of the video is simple and straightforward, focusing on the girl and her expressions. The lighting is soft and even, suggesting an indoor setting. The video does not contain any additional objects or elements that would provide context or additional information about the girl or her surroundings."
+    ],
+    "video_ids": [
+        "VU9WgOSCiKc_18_0to170"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young girl with dark hair, wearing a colorful, striped tank top.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video includes a young girl with dark hair wearing a colorful, striped tank top. The description matches the first girl shown in the video, who is wearing a blue and white striped dress with ruffles, which can be considered a colorful, striped tank top. The presence of other children does not contradict the core description as they are additional elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young girl with dark hair wearing a colorful, striped tank top. The tank matches the visual girl in in the video, who is wearing a rainbow, pink striped tank. pinkuffles. which can be considered a type, striped tank top. The second of another elements in not contradict the core description.\"\n the are also elements in}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3d59866091bfb17c2a9f0319f969b546c7f3dd05
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c8990c447d3a00cf9785a96154fee6d47b3fd33aaa1d1b100b1ce3ae0b0072b
+size 82416
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..036ee79db1c29f814789c6e7d17f1603c580c1f3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a journey on a winding mountain road. The scene is set against a backdrop of a clear blue sky and a rocky cliff. The road is marked by a yellow warning sign that cautions drivers to slow down ahead.   In the first frame, a white truck is seen driving on the left side of the road. The truck is moving away from the camera, giving a sense of depth and distance.   In the second frame, the same white truck is now closer to the camera, indicating that it has traveled further down the road. The truck is still on the left side of the road, adhering to the traffic rules.  In the third frame, the white truck has now passed the yellow warning sign. The sign is still visible in the background, serving as a reminder of the road's challenging nature. The truck is still on the left side of the road, maintaining its position.  Throughout the video, the white truck is the main subject, with the winding mountain road and the rocky cliff serving as the backdrop. The yellow warning sign adds an element of caution and anticipation to the journey. The clear blue sky adds a sense of calm and serenity to the overall scene."
+    ],
+    "video_ids": [
+        "WbPT_4rJJ2g_8_22to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white truck, a yellow warning sign? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a white truck driving on the road and a yellow warning sign that reads 'WARNING SHARP CURVES AHEAD SLOW DOWN NOW'. These objects are prominently featured and match the description exactly.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a white truck driving on a road, a yellow warning sign is is 'STOP'ARP CURVE'HEAD'LOW DOWN'' The elements are present displayed in match the description provided.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..26031a723437c11cd816c864a7db9fa38e585bdd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d0b3613625caed165fa58cfd33a60f5a0b2e4dae6d89bc907a815478d69d756
+size 294283
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..298b5b5fbf6e8850d48c7bc9d4207299414ed965
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man sitting in a red chair, wearing a blue shirt and a gray vest. He has a mustache and is smiling. The man is gesturing with his hands as he speaks, indicating that he is engaged in a conversation or giving a presentation. The background is black, which contrasts with the red chair and the man's colorful attire. The overall style of the video is simple and straightforward, focusing on the man and his actions."
+    ],
+    "video_ids": [
+        "57tLwFCvsvE_3_0to140"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man sitting in a red chair, wearing a blue shirt and a gray vest. He has a mustache and is smiling.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting in a red chair, wearing a blue shirt and a gray vest, with a mustache and a smiling expression. These elements match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting in a red chair, wearing a blue shirt and a gray vest. which a mustache and a smile expression. The elements match the description provided, the questionObject(s)' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9f2f794ba84c9212657d1847456e4fc4e8016dd5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0cd23d9898d33dff915fb701904f63ef3b907f5262d834a2e4b872af24a23f91
+size 167624
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..5a496bb1cafd1d3b7851b8a1769708ea9900416b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a colorful and vibrant depiction of a religious scene, featuring a large golden Buddha statue at the center, surrounded by smaller golden statues of monks and other figures. The statues are arranged in a semi-circle around the Buddha, with each figure holding a different pose and object, such as a fan, a book, or a lotus flower. The background is a painted mural of a serene landscape with trees, mountains, and a river, adding to the peaceful and spiritual atmosphere of the scene. The style of the video is artistic and detailed, with a focus on the intricate craftsmanship of the statues and the rich colors of the mural. The video captures the essence of the religious narrative and the cultural significance of the figures, making it a visually appealing and engaging piece of art."
+    ],
+    "video_ids": [
+        "3Ok6K80kWb8_17_478to648"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large golden Buddha statue, smaller golden statues of monks and other figures, and a painted mural of the landscape.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a large golden Buddha statue seated on a lotus, surrounded by smaller golden statues of monks and other figures. Behind them is a painted mural depicting a landscape with trees and mountains, which matches the described 'Object(s)' condition. The presence of logos and social media icons does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a large golden Buddha statue at in a lotus pedestal surrounded by smaller golden statues of monks and other figures. The the, a painted mural depicting a landscape with trees, mountains, which ful the description 'p(s)' condition.\"\n The video of the in the media icons in not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..09ffd68a2e55f3567897ccc7c61525d70eee9953
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5bbf344d25cf579cb1714d707c143695988afdfcbbe1c98bb2b23648f474fb9c
+size 215657
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..1701c06c524167bec2a955324a852d2672b3f0d6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person with manicured nails and a ring on their finger, using a sewing machine to work on a quilt with a floral pattern. The person is focused on their task, carefully guiding the sewing machine needle through the fabric. The sewing machine is white and has a needle that is currently in use. The quilt is placed on a table, and the person is seated in front of it, working diligently on their craft. The overall style of the video is a close-up, real-life depiction of a person engaged in a sewing project."
+    ],
+    "video_ids": [
+        "2u8AT6lJK_c_53_0to123"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person, sewing machine, quilt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person operating a sewing machine while working on a quilt. The hands of the person are visible, guiding the fabric under the needle of the sewing machine, and the quilt, composed of various patterned fabric squares, is being sewn together. All core elements mentioned in the condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a person using a sewing machine, working on a piece. The person of a person are visible, and the fabric through the needle of the sewing machine, which the quilt with which of colorful colorfuled fabrics pieces, is being sewn.. The elements elements of in the condition are present and accurately depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9dad31ac16b5cfd418e59e93e9d4ef9a66546d74
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c6f870410e627829316d0e1b8c2838d8c7734e6069a78a51a5ce66ad23e68be
+size 332611
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..78153928be97ab0653102bff6e7ea97ba281ce01
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of interaction between two individuals, possibly politicians, in a public setting. The man, wearing a striped shirt, extends his hand towards the woman, who is dressed in a blue jacket. They are surrounded by a crowd of people, some of whom are holding papers, suggesting a formal event or gathering. The atmosphere appears to be cordial and respectful, with the man and woman engaging in a handshake. The style of the video is candid and unposed, capturing a genuine moment of interaction between the two individuals."
+    ],
+    "video_ids": [
+        "1flGkxGYzgM_17_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two individuals, a man wearing a striped shirt and a woman dressed in a blue jacket, and a crowd of people.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a striped shirt and a woman in a blue jacket interacting at a podium, surrounded by a crowd of people, which matches the 'Object(s)' condition described. The core elements are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts two man wearing a striped shirt and a woman in a blue jacket shaking with what public in which by a crowd of people. which align the descriptionObject(s)' condition described. The presence elements of present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..44e2238f694f453935d0adc156de4e074110e62b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e462adfe212efb5a4ce50ef113ca1d3c6208438b28fbc24743f349178777dfe6
+size 226992
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..f806fc6f0cbfe0aca425fd288cd063fc2754d79b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young woman is seen cutting a cake with a knife. The cake is white and has the words \"Happy Birthday\" written on it in chocolate. The woman is wearing a white shirt and is standing at a table with a black and white tablecloth. The tablecloth has a floral pattern. The woman is using a knife to cut the cake. The cake is placed on the table. The woman is the only person in the video. The video is a simple, everyday scene."
+    ],
+    "video_ids": [
+        "QknT67Jjzsg_10_44to172"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman, a white cake with 'Happy Birthday' written in chocolate, a knife.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman cutting a white cake with 'Happy Birthday' written in chocolate icing using a knife. All core elements from the description are present and accurately depicted without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman cutting a white cake with 'Happy Birthday' written in chocolate.. a knife. The the elements of the description are present: accurately depicted in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..132af460f376b9fb3276719a5e83337d0868e600
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c4fefd5fa5efa4dfd5fe74e0021dfe38cd4213015e8cebbc1be3ab6c326ae104
+size 227274
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..56cd2fc1817fc90ecf59103d30d80e8802c48c38
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a luxury car, focusing on the steering wheel and dashboard. The steering wheel is prominently displayed, featuring the distinctive logo of the car manufacturer. The dashboard is sleek and modern, with various controls and displays. The car's interior is well-lit, highlighting the high-quality materials and craftsmanship. The video is shot in a close-up style, focusing on the details of the steering wheel and dashboard. The overall impression is one of luxury and sophistication."
+    ],
+    "video_ids": [
+        "jS_2BaTQhs8_28_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, dashboard? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a close-up of the steering wheel with the Lamborghini logo, which satisfies the 'Steering wheel' condition. The dashboard is partially visible in the background, fulfilling the 'dashboard' condition. No elements contradict these core descriptions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a steering-up view a steering wheel and a logoorghini logo, which is the 'Steering wheel' condition. Additionally dashboard is also visible in the background, showing the 'dashboard' condition. The additional contradict the descriptions descriptions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a5e376020155632943ef92a1923e35d25091414b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1214-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13beca0aa2bad13ea8be7c5ad1c66e8393d43d7e5001aba71f428dcd3f70d01b
+size 172422
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..b299217c791a5b67df9f76b262ddc1a378eb120a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a fashion show taking place on a runway. The central figure is a model striding confidently down the runway, wearing a beige jumpsuit and a matching beige coat. The jumpsuit is form-fitting, accentuating the model's figure, and the coat is draped over one shoulder, adding a touch of elegance. The model's face is serious, reflecting the intensity of the moment.  In the background, other models can be seen walking in the same direction, each showcasing their own unique style. The runway is set against a backdrop of lush greenery, providing a stark contrast to the urban fashion scene. The overall style of the video is dynamic and vibrant, capturing the energy and excitement of a live fashion show."
+    ],
+    "video_ids": [
+        "DfxMSMm0BSQ_2_50to186"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A model in a beige jumpsuit and coat, other models walking in the same direction, and the runway.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a model in a beige ribbed bodysuit with a beige coat draped over her shoulders, walking confidently down a runway. Other models are visible in the background, also walking in the same direction, and the runway is clearly depicted as a light-colored path extending into the distance. The scene matches the described 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a model in a beige jumpsbed jumpsysuit and a matching coat, over her shoulders, walking down down a runway. There models in also in the background, also walking in the same direction. which the setting setting clearly defined with a straight-colored, surrounded into the distance. The setting is the description elementsObject(s)' condition.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f1848cad73ae3a0e08262a3144d499887113f3bf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8843577a42d6562d605407262599c338ca85784c34a5d6553cb8de979eb4be76
+size 231935
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..2c0a14c70344e0a493d5d7ffc99945dd696d0491
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young man and woman are standing in front of a green sports car, smiling and looking at the camera. The car is parked on a grassy area with trees in the background. The man is wearing a white shirt with black designs, while the woman is wearing a black shirt. The car has a sleek design with a black roof and black rims. The background features a clear sky and the trees are in full bloom, indicating that it is likely spring or summer. The overall style of the video is casual and friendly, with the two individuals appearing to be enjoying their time together."
+    ],
+    "video_ids": [
+        "B3y-t8B6jPs_96_0to191"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man and woman, a green sports car with a black roof and black rims, a white shirt with black designs, a black shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young man and woman standing in front of a green sports car with a black roof and black rims, which matches the description. The man is wearing a white shirt with black designs, and the woman is wearing a black shirt. All specified objects are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man and woman standing next front of a green sports car with a black roof and black rims. which matches the description. The man is wearing a white shirt with black designs, and the woman is wearing a black shirt. The elements elements and present and accurately depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..99de30d77164669c1e8189fc5c67a96809f67233
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1673c9198d412d40055d5e3797ed8a3785ef2548097552317c4abc5a403106dd
+size 199563
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..45c9889eba6906118b925c41faf07d2b19cee670
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a steering wheel in a car, with the camera moving from the left side of the wheel to the right side. The steering wheel is black and has a red and white logo in the center. The car's dashboard is visible in the background, with various controls and gauges. The steering wheel is made of leather and has a shiny appearance. The car's interior is red, and the steering wheel is positioned in the center of the dashboard. The camera moves smoothly, providing a clear view of the steering wheel and its surroundings. The video is shot in a realistic style, with no filters or special effects."
+    ],
+    "video_ids": [
+        "S5S_iAvnD2g_14_40to203"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, car's dashboard, controls and gauges? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a steering wheel with a Corvette logo, part of the car's dashboard, and various controls and gauges. The elements described in the condition are all present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a steering wheel, a central logo, a of a car's dashboard with and various controls and gauges. The presence in in the question are present present and accurately depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0998304e2a5ee5f3903f6ca65ca3821567f920f2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2350d194f19ef6f542f5199217276ad514fd0a4220d8d014ba7c709bd6664251
+size 178955
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..065d7544622e96d575429b408d1c9b519ba70193
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a scene at a gas station on a cloudy day. The gas station sign displays the prices for unleaded and premium gasoline. The first frame shows the sign with the price of unleaded at $2.02 per gallon. In the second frame, the price of premium gasoline is shown as $2.42 per gallon. The third frame shows a car driving past the gas station. The car is a silver sedan. The gas station is located on a street with a sidewalk and a parking lot. The sky is overcast, and the ground is wet, suggesting that it has recently rained. The overall style of the video is a straightforward, unembellished documentation of a commonplace scene."
+    ],
+    "video_ids": [
+        "PHqVAczqzgo_0_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Gas station sign, car (silver sedan)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a gas station sign displaying prices for 'Unlead' and 'Prem' fuel, and a silver sedan is visible in the parking area. These elements match the requested objects in the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a gas station sign with the and gasGasle' and 'Unier gas, with a silver sedan driving visible on the background area. The elements match the description ', the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..31ca3be1537bacc0aa48e1191ddc4e0ed0695606
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea6024eaf8174500f8ada2e938dd7bc8d21a6cefac304f6b595cbc1be6f82650
+size 105342
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..efba77519fa87ad1dcc0c3b4cc38771decb6a9ba
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a pile of almonds, with the focus on the texture and color of the almonds. The almonds are in various stages of being cracked open, revealing the inner nut. The style of the video is a close-up, still-life shot with a shallow depth of field, emphasizing the texture and color of the almonds. The lighting is soft and even, highlighting the natural colors of the almonds. The video does not contain any text or additional elements. The overall impression is one of simplicity and focus on the subject matter."
+    ],
+    "video_ids": [
+        "-5jSJ_6W1DY_14_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Pile of almonds in various stages of being cracked open.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a pile of nuts that appear to be almonds, with many in various stages of being cracked open. The shells are visibly split or broken, matching the description. There are no conflicting elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a pile-up of a pile of almonds, appear to be almonds, with some of their stages of being cracked open. The focus are partially broken, partially, revealing the description of The are also additional elements in contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2bbf76f209964d13b36f0c6d3ba3532c6dc52750
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5cb9790ba76fcd714e88324c15bf642c0e5f4d08dd0218c45e8d447f4a27ce2
+size 78426
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..453f62a18d6bb4ec18d2665241e49f69bf2e98d8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a sequence where a row of small yellow rubber ducks is lined up on a textured asphalt surface. Each duck has a distinct facial expression, ranging from neutral to slightly angry or surprised. The scene is set under a vehicle, with the edge of the car's tire visible at the top right corner of the frame. As the video progresses, the camera pans slowly to the left, revealing more of the ducks and the surrounding area. The ducks remain stationary throughout the video, maintaining their positions relative to each other. The lighting suggests an outdoor setting during daylight hours, with shadows cast by the ducks indicating the presence of sunlight. The overall atmosphere is playful and whimsical, with the ducks' expressions adding a touch of humor and personality to the scene."
+    ],
+    "video_ids": [
+        "a6899a8f6a92bcc93594103122e08e20b72a4056876d81b0c9556c50774cf7de"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Row of small yellow rubber ducks with distinct facial expressions? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a row of small yellow rubber ducks, each with distinct facial expressions (some with winking eyes, others with neutral or slightly angry expressions). These ducks are the primary focus and are consistently visible throughout the video, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a row of small yellow rubber ducks with each with a facial expressions,open with openinking eyes and others with open or slightly open expressions). The ducks are positioned primary focus and are arranged positioned throughout the frames, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a172fc17669e3e56c3324cb09b22be8ee8941870
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf40e32265fb3cc05dba1ecd5ec3204f463a4c97f8582e59c7ad6c49d597ef5d
+size 118087
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..a77f977a26339241632695a7d3d753521047921e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene by a calm body of water, likely a river or pond, surrounded by lush greenery. The water is still, reflecting the trees and sky above, creating a mirror-like effect. A prominent feature in the foreground is a fallen tree trunk extending diagonally across the frame from the bottom left to the top right. This trunk appears weathered and partially submerged in the water.  As the video progresses, a small animal, possibly a beaver or otter, emerges from beneath the water near the base of the tree trunk. The creature's fur is dark and wet, and it moves quickly, its tail flicking rapidly as it climbs up the trunk. The animal's movements are agile and purposeful, suggesting it might be foraging or exploring its surroundings. The background remains consistent throughout, with the tranquil water and the reflections of the trees providing a peaceful backdrop to the animal's activity. The overall atmosphere of the video is one of quiet natural beauty, with the animal"
+    ],
+    "video_ids": [
+        "7793bd2b65b98ec7ffafe07da4a5fddece1700dd92def01a7d8db7e3c1865b31"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A fallen tree trunk and a small animal (possibly a beaver or otter).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a fallen tree trunk extending across a body of water, and a small animal with a bushy tail, consistent with a beaver or otter, is visible near the trunk. The animal's movements and position align with the description, and no elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a fallen tree trunk in into a body of water. and a small animal, a darky tail, which with the beaver or otter, is seen. the trunk. The animal is behavior and the align with the description of and there additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5b815ba80ee45687ac3d939744235a7cbb310f50
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:69990f039d16753f2f5c0cbed36f177e937f1867623d7a72f6cffc60f2324b26
+size 274434
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..eb82c11f709c7b22e59ed27219af5e5ba1960fd3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a baseball game. The main focus is on a batter, dressed in a white uniform with the number 99, who is in the midst of swinging his bat. The batter's stance and the motion of the bat suggest a powerful swing. Behind the batter, a catcher and an umpire are positioned, ready to react to the batter's swing. The catcher is wearing a baseball glove, ready to catch the ball if the batter misses. The umpire is attentively observing the play, ready to make a call. The background of the video shows a baseball field with a green outfield and a brown dirt infield. The field is surrounded by a fence, and there are spectators watching the game from the stands. The video is shot from a side angle, providing a clear view of the batter, the catcher, and the umpire. The style of the video is realistic, capturing the intensity and excitement of a live baseball game."
+    ],
+    "video_ids": [
+        "NeMrpgtFNHY_58_0to173"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Batter, catcher, umpire? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a batter (wearing number 99), a catcher in full gear, and an umpire positioned behind the catcher, all of whom are central to a baseball game scenario. These three figures are accurately depicted in their respective roles, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a batter,wearing a 99), a catcher, a gear, and an umpire positioned behind the catcher. all of which are typical to the baseball game.. The elements elements are the represented as the respective roles, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c42835df186c09d3a66bbd17859c161fc9a618fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1215-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:16806e067a901387c6c0022b2d1e0eb1bab5e28325d1b851a2f4af590b116c6b
+size 249002
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..7c0d3966577178c906e89f2842ed0be78b7e8d75
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit sitting in a well-lit room with a rustic design. He is gesturing with his hands, suggesting he is in the middle of a conversation or explanation. The room has a warm and inviting atmosphere, with wooden walls and a comfortable couch in the background. The man appears to be the main subject of the video, and his actions and expressions are the focal point. The overall style of the video is professional and polished, with a focus on the man's speech and the room's ambiance."
+    ],
+    "video_ids": [
+        "PWoCR0aSH-M_9_0to178"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a suit jacket over a collared shirt, which matches the description of 'A man in a suit'. The setting and additional elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed a suit,, a dressared shirt, which align the description of aA man in a suit'. The presence and attire elements, not contradict this description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..851dab161d7a0d39c09a68ea9ed180ca53800a9b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:126ced121b5f738ef9b4f460b92b42ae9c4a56890553933a5092e2a29b77d4c2
+size 140184
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..d0a8f1761ffd2f2c20a57387e59d5aab7965fe64
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen interacting with a silver Porsche sports car. The car is parked on a road, and the man is positioned behind it, leaning on the car's rear. He is wearing a white sweater and blue jeans, and he appears to be examining the car's license plate, which reads \"RK19 FSO\". The car itself is a sleek, silver Porsche, and it features a distinctive black rear wing. The setting is a forested area, with trees visible in the background. The overall style of the video is casual and relaxed, with a focus on the man and the car."
+    ],
+    "video_ids": [
+        "OoIlk4RkS28_23_93to229"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a silver Porsche sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man crouching next to a silver Porsche sports car, specifically a 718 Boxster GTS, as described. The core elements of the request \u2014 a man and a silver Porsche sports car \u2014 are accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man interactingrouching next to a silver Porsche sports car. which a Porsche918 modelster,TS, as indicated. The car elements of the '\u2014 a man and a silver Porsche sports car \u2014 are present represented in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cc7f904494590252afe1b79e0b8399c55540e48b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d90c6bde85e2041a8be7044b8364930c19ba7ba0c85d9e01c7ba044a759de388
+size 157904
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..40a11e887eb21638b6fa15ce3889792108c60ccc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen holding a red packet of Pikopiko over a small yellow cup with a cartoon character on it. The cup is placed on a rock near a body of water. The person appears to be in the process of pouring the contents of the packet into the cup. The overall style of the video is casual and informal, with a focus on the action of pouring the packet into the cup. The setting appears to be outdoors, possibly near a river or a lake, adding a natural element to the scene. The video captures a simple, everyday moment with a sense of playfulness and enjoyment."
+    ],
+    "video_ids": [
+        "6AJRlW-9LEo_21_0to135"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person, red packet of Pikopiko, small yellow cup with a cartoon character, rock, body of water? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a person's hand holding a red Pikopiko packet, pouring its contents into a small yellow cup featuring a cartoon character. The cup is placed on a rock, and a body of water is visible in the background. All specified objects are present and correctly depicted without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a person pouring hand holding a red packetopiko packet, pouring its contents into a small yellow cup with a cartoon character. The cup is placed on a rock, and the body of water is visible in the background. The the objects are present and correctly depicted in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5fe56eeec294fa413842a42171f604baf882025c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d203bf74287e352affaca2c7249bc0684c9c28b6f587894bc1fc74409cb1037
+size 263030
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..d21894b4bef9710442e359c17f0e3edb348fb694
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is an aerial view of a cityscape, showcasing a tall, white skyscraper as the focal point. The skyscraper is surrounded by other buildings of varying heights, creating a dense urban environment. The city is lush with green trees and grass, providing a stark contrast to the concrete structures. The sky is clear and blue, indicating a sunny day. The video captures the essence of a bustling city, with the towering skyscraper standing as a testament to human achievement amidst the natural beauty of the surrounding greenery."
+    ],
+    "video_ids": [
+        "FtxXJkQdMXo_512_87to220"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Tall, white skyscraper, other buildings, green trees, grass? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a tall, white skyscraper as the central object, surrounded by other buildings of varying heights and designs. Green trees and patches of grass are visible throughout the urban landscape, including around the base of the skyscraper and in adjacent areas. These elements align with the described conditions without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a tall, white skyscraper as the central object. surrounded by other buildings of varying heights and colors. The trees and grass of grass are visible throughout the scene landscape, fulfilling in the skys of the skyscraper and in the areas. The elements collectively with the description ', any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b79171bc90023b9ecbd58722ef4cc7d494c2bd7a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1dc05ce8d1a2b2fa5142623df6197fcecb86a25dd59589af6e5b411729725646
+size 325259
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..ee6583c59b87f9518e5fb6bbf877a3673875712d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men in a kitchen, preparing food. The first man is holding a bowl and pouring something into a pot on the stove. The second man is pouring something from a measuring cup into a mixing bowl. The kitchen is well-equipped with various utensils and ingredients. The men are focused on their task, and the atmosphere is one of concentration and collaboration. The video is likely a cooking tutorial or a demonstration of a recipe."
+    ],
+    "video_ids": [
+        "1IFSMmdqMiU_13_77to222"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, a bowl, a pot, a stove, a measuring cup, a mixing bowl.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men in a kitchen setting. One man is pouring from a measuring cup into a mixing bowl, while the other is holding a small plate. A pot is visible on the stove in front of them. All the specified objects \u2014 two men, a bowl, a pot, a stove, a measuring cup, and a mixing bowl \u2014 are present and clearly identifiable in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two men in a kitchen setting. One man is holding a a measuring cup into a pot bowl, which the other man holding a bowl bowl. There pot is on on the stove, the of them. The the objects objects ( two men, a bowl, a pot, a stove, and measuring cup, and a mixing bowl \u2014 are present and being identifiable in the scene.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f87d1d8102418557e4bf69bf2bb4e18a6c706409
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1daf8d773276e6d47f380a919633e00d7fa2be9beb26e33747a38f58bb43bf11
+size 168974
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..38a83b2a5c16355bacea656ab4b3bb71e96e2fad
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a group of people walking along a narrow, winding path that hugs the side of a steep cliff. The cliff is made of red rock and is quite steep, with the path appearing to be carved into the side of the cliff. The path is narrow and winding, with the cliff on one side and a steep drop-off on the other. The group of people is spread out along the path, some walking ahead and others following behind. The path is lined with a stone wall on one side, providing a sense of security for the walkers. The video is shot from a high angle, looking down on the path and the people walking along it. The overall style of the video is dramatic and adventurous, capturing the thrill of walking along a cliff-side path."
+    ],
+    "video_ids": [
+        "8WLi9l-qr3g_36_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A group of people? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a group of people standing along a narrow path on the side of a cliff. They are visible throughout the frames, confirming the presence of a group of people as described in the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts shows a group of people walking on a pathway pathway carved a side of a mountain. This are positioned and the frames, indicating the presence of the group of people as described.\"\n the condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..265d9b972c3f67377c64c149b41582f434a3e76b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5026b4ba77160843ac738c3158971833c3f55def62f872bc0c82f32a67961d1c
+size 329876
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..a28e252f9d1a12c60bded1f655bfe3eb0c520c9e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a brown SUV parked on a concrete surface near a body of water, possibly a lake or a river. The car is positioned facing the water, with its rear end visible to the viewer. The sky is partly cloudy, suggesting a calm and serene day. The water reflects the sky and the surrounding landscape, adding to the tranquil atmosphere. The concrete surface on which the car is parked appears to be a small parking area or a roadside stop. The overall style of the video is realistic and naturalistic, capturing the beauty of the outdoor setting without any artificial embellishments. The focus is on the car and its immediate surroundings, with no additional elements or distractions. The video likely serves as a promotional or advertising material for the SUV, highlighting its design and features in a picturesque and appealing environment."
+    ],
+    "video_ids": [
+        "2PpO48y-Tk8_0_0to144"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A brown SUV? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a brown SUV parked by the water, which matches the description. The vehicle is clearly visible and occupies the central focus of the scene, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a brown SUV, on a side, which align the description of The vehicle's the visible and the a central part of the frames, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5fca351a6b3f3e2dc0d4945af0019f82971943ef
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec7b63033ffb6c7e01d22b7cf0f98c86a84e319320f4c14ee34254823bd91811
+size 146062
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..1b5e84ac62cc9ba0273b8ff6b61e8bdf12fa6b81
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red sports car parked on the side of a street. The car has its trunk open, revealing a black interior. The car is parked in front of a beige building with a red roof. The street is lined with palm trees and there is a white car parked on the opposite side of the street. The car is facing the camera and the trunk is open. The car is parked on the side of the street. The car is red. The car is parked in front of a building. The car is parked on the street. The car is parked on the side of the street. The car is parked on the street. The car is parked on the side of the street. The car is parked on the street. The car is parked on the side of the street. The car is parked on the street. The car is parked on the side of the street. The car is parked on the street. The car is parked on the side of the street. The car is parked on the street. The car is parked on the side of the street. The car is parked on the street. The car is parked on the side of the street. The car is parked on the street. The car is parked on the side of the street. The car is parked on the street. The car is parked on the side of the street. The car is parked on the street. The car is parked on the side of the street. The car is parked on the"
+    ],
+    "video_ids": [
+        "5Sf-yrM5LEM_7_0to134"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red sports car and a white car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red sports car parked on a street, with its rear door open. A white car is also visible in the background, parked further down the road. Both objects are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a red sports car with on the street with with its hood hood open. Additionally white car is also visible in the background, parked further down the street. The elements match present and match with the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..95f215ae9365753eb9aba8938bf62b690e413a15
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1216-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:82942a3ea9b78b7cdd328edd31358677140cb439a7a541ae218e40a6f448c17b
+size 178975
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..36ca387979a464efdd75a2d50f6c28ee1b94bd94
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen preparing a cake batter using a red mixer. The person is wearing a green apron and is using a wooden spoon to mix the batter in a glass bowl. The mixer is placed on a kitchen counter, and there is a spatula nearby. The person is carefully stirring the batter, ensuring that it is well mixed. The scene is set in a kitchen, and the person appears to be focused on their task. The video captures the process of preparing a cake batter, from mixing the ingredients to using the mixer to blend them together."
+    ],
+    "video_ids": [
+        "VEQ-EQcDejA_25_28to221"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a red mixer, a wooden spoon, a glass bowl, and a spatula.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person wearing a green apron, using a red stand mixer with a glass bowl attached. The person is also using a wooden spoon and a spatula to scrape down the sides of the bowl. All the specified objects are present and correctly identified in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a person using a green apron, a a red mixer mixer with a wooden bowl attached. A person is using using a wooden spoon to a spatula to mix the the sides of the bowl. The the objects objects are present and used used in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..072507743bafe6f1e044e41a462f561005146875
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:607781a55715b89701fc91d029b0661190cfd3f95cf38160b52f162ba7ea5a30
+size 271817
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..ca2d04b1c8f30fbaa70aa04825e8d75cf4ff1fa6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen giving a speech at a conference. He is dressed in a black sweater and is standing behind a podium. The man is gesturing with his hands as he speaks, indicating that he is in the middle of his speech. The background of the video features a banner with the word \"Wondercon\" written on it, suggesting that the conference is related to Wondercon. The man appears to be the main speaker at the event, as he is the only person visible in the video. The overall style of the video is straightforward and informative, focusing on the man's speech and the context of the conference."
+    ],
+    "video_ids": [
+        "C6yXQyaxQ4E_30_34to190"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man dressed in a black sweater standing behind a podium.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man standing behind a podium, wearing a black sweater (specifically, a black zip-up hoodie over a blue sweater). The core description is accurately fulfilled, with no significant contradictions. Additional elements like the 'WONDERCON' backdrop and the man's raised hand do not conflict with the primary object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man standing behind a podium, wearing a black sweater.whichally, a black button-up sweater). a black shirt). He setting elements of largely represented, with the significant contradictions or The elements like the backgroundbooDER'' sign do the microphone's gestures hand gestures not conflict with the main description description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..38edf8fc93f0bb76921706e751833b3caf2b423d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8dfd7637833d0c7f367511a7bd9e5f9016a59ac2a535e9954bf6442c722cf043
+size 161021
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..4b9acbfa7a6bbe6b2d8940547e1e1eaadf54c70d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red sports car and a black Jeep parked side by side on a concrete driveway. The car is sleek and shiny, with a large front grille and a prominent hood scoop. The Jeep is rugged and robust, with a large front bumper and a spare tire mounted on the back. Both vehicles are parked in front of a garage door, which is closed. The garage door is made of metal and has a white frame. The driveway is made of concrete and is clean and well-maintained. The sky is clear and blue, suggesting a sunny day. The video is a simple, straightforward shot of two vehicles parked side by side, with no action or movement. The style of the video is realistic and unembellished, focusing on the vehicles and their immediate surroundings."
+    ],
+    "video_ids": [
+        "_kQhXeC9lrE_8_116to269"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red sports car and black Jeep? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red sports car (a Dodge Challenger SRT) and a black Jeep (a Jeep Wrangler) parked side by side in front of a garage. These objects match the description exactly, with no contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a red sports car andlikely Honda V)RT Demon and a black Jeep (a Jeep Renangler) parked in by side in front of a garage. The are match the description provided, and no additional or}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..22186e3e26ccee7e4dfe8cb15a54ea1ca8c2d055
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b3d59fdf3b078c6ed8c80db10d4cc2f364bb732f21ad512bf9c4d9d7074f1ffc
+size 161268
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..ed91910e0fb4827ff23fb091952fcd2f55576b25
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard, wearing a pink shirt, sitting on a couch in a colorful room. He is holding a bag of chips and a drink. The room is decorated with various objects, including a TV, a clock, and a window. The man appears to be relaxed and enjoying his snack. The overall style of the video is casual and laid-back, with a focus on the man's leisurely activity."
+    ],
+    "video_ids": [
+        "LSDvpgQd5mo_122_0to103"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard wearing a pink shirt, sitting on a couch, holding a bag of chips and a drink.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard wearing a pink shirt, sitting on a couch. He is holding a bag of chips (visible in front of him) and a drink (a water bottle is visible to his left). The presence of a green diamond graphic does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a man with a beard wearing a pink shirt, sitting on a couch. He is holding a bag of chips andspecific in the of him) and appears drink (not straw bottle is partially in his side). The setting of the TV plant-shaped in not contradict the description description and}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fbe8d089b7204492965e53f15c4f54ea67f6dee2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc06f1adb88a46db5c7a79280a51a1c580c7c1c723b1b8190e4e69201d3c52cd
+size 111569
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..dea0f9d9d3cf386847138356f1c86affc14d26db
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man sitting at a desk in a room with a computer monitor in front of him. The monitor displays a ship on the screen. The man is wearing a blue jacket with a red and white design on the front. He has short hair and is looking directly at the camera. The room has a white wall with a floral pattern and a shelf above the desk. There is a bottle on the shelf and a keyboard on the desk. The style of the video is a simple, straightforward shot of the man at his desk, with no additional action or movement."
+    ],
+    "video_ids": [
+        "4kAFtB9w3Yg_149_0to164"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a computer monitor, a blue jacket, short hair, a bottle, and a keyboard.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with short hair wearing a blue jacket, sitting in front of a computer monitor. A bottle is visible on the shelf behind the monitor, and a keyboard is on the desk. All specified objects are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing short hair wearing a blue jacket, sitting in front of a computer monitor. There keyboard is visible on the shelf behind him man, and a keyboard is on the desk in The the elements are present and match with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..240004443529d235866e92f12dfcee7905b4be07
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2cc6833f35b4307274cadc81ed93c2c3cf942a75429642de1399dc37e33f7f52
+size 103484
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..991cd45d94cedec0960b4bf0112c86f8414475da
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a workshop, wearing a blue shirt and glasses, with a beard and mustache. He is holding a wooden object in his hands. The workshop is equipped with various tools and equipment, including a drill press and a vice. The man appears to be in the process of working on the wooden object. The workshop has a patriotic theme, with an American flag hanging on the wall. The man seems to be focused on his work, and the overall atmosphere of the video is one of concentration and craftsmanship."
+    ],
+    "video_ids": [
+        "ZHr8uqUT4bA_0_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue shirt and glasses, a wooden object, a drill press, a vice? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue shirt and glasses, holding a wooden object. In the background, there is a drill press and a vice visible, matching the described objects. The scene is consistent with the 'Object(s)' condition without contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue shirt and glasses, holding a wooden object, The the background, there is a drill press and a vice,. which the description elements. The presence appears set with the givenObject(s)' condition provided any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..92c3bf4c198d90dd6a2aebac85bad4b4da9fdb31
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7ccb40e7ad33d77c55e617beb9d588e63250c2b70775636b7e3399149c8c8604
+size 166055
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..4fd3ba7cf5c7ec70801a45802e3d26ccb530a812
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a basketball game. The main focus is on a player in a white jersey, who is in the process of dribbling the ball. He is being closely guarded by two players in black jerseys, who are attempting to block his path. The player in white is making a determined effort to maintain control of the ball, while the players in black are using their bodies to create a physical barrier. The intensity of the game is palpable, with the players' movements and expressions conveying a sense of urgency and competition. The background is filled with the blurred figures of other players and spectators, adding to the sense of a bustling, high-stakes environment. The video is shot from a low angle, emphasizing the players' movements and the action on the court."
+    ],
+    "video_ids": [
+        "Nsd4Zjj8eJY_1_0to126"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a white jersey, two players in black jerseys, other players, spectators? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a white jersey holding a basketball while being closely guarded by two players in black jerseys. Additional players and spectators are visible in the blurred background, fulfilling the described conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a basketball in a white jersey, a basketball, being guarded guarded by two players in black jerseys. There players and spectators are visible in the background background, which the ' conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9504bb3f65c6515e884749e4de25b1ebab5b37e0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1f2cd6ef10e83276addf5d4414ae6aff76549d585b07aa2c53d579d0910693d6
+size 314012
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..0c1f33629dbc92b1224a0359e6617e41b03abbf3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and glasses, wearing a blue polo shirt. He is seated outdoors, surrounded by lush greenery and a stone wall. In the first frame, he is seen with a slight smile, looking directly at the camera. In the second frame, he raises his index finger, indicating a point of interest or emphasis. In the third frame, he continues to hold up his index finger, maintaining eye contact with the camera. The style of the video is a straightforward interview or discussion, with the man appearing to be engaged in a conversation or making a statement. The setting is serene and natural, providing a calm and peaceful backdrop for the man's speech."
+    ],
+    "video_ids": [
+        "-XJzj_VwiR8_36_113to237"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and glasses, wearing a blue polo shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and glasses, wearing a blue polo shirt, which matches the description. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man with a beard and glasses, wearing a blue polo shirt. which matches the description provided The man appears the elements in not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5b07e5515abc38fee4d358b7b18868c2fe14652a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1217-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75bc7c5fe4024d57bf7ac51f66edde0c001a50a19143537c0804d157bf3a6a56
+size 209645
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..9bd78af2491e7f74c6f7b9ed896a87ebfb98d241
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen enjoying a beverage from a white cup. The cup is filled with a light brown liquid, possibly a coffee or tea. The person is holding the cup with one hand, while the other hand is holding a piece of food, which appears to be a pastry or cookie. The food is being dipped into the beverage, suggesting that the person is enjoying a snack along with their drink. The setting appears to be casual and relaxed, with the person seated comfortably. The focus of the video is on the person and their actions, with the background being less prominent. The style of the video is simple and straightforward, capturing a common everyday moment."
+    ],
+    "video_ids": [
+        "IE7_SkkzKwA_27_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person, white cup, light brown liquid, piece of food (pastry or cookie)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person holding a white cup containing a light brown liquid, and dipping a piece of food (which appears to be a pastry or cookie) into it. All core elements described in the condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a person holding a white cup filled a light brown liquid, which the a piece of food (which appears to be a pastry or cookie) into the. The the elements of in the question are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d9434a0e693b135853a53b55c92a8f28ccbc5bfe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:39058d3295ae0fd3a964e165a2165a4f0e87fa1050a82a5483e680f10b815fd4
+size 113745
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..741fc53cdcf6a6e001a1569eaedb2b852d57abcd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman with blonde hair is seen sitting at a dining table in a kitchen. She is wearing a pink scarf and a pink shirt. She is holding a magazine and appears to be reading it. The kitchen is well-lit and has a refrigerator in the background. The woman seems to be enjoying her time alone, immersed in the content of the magazine. The overall atmosphere of the video is calm and peaceful."
+    ],
+    "video_ids": [
+        "7vmEbSCHOt4_13_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with blonde hair, a pink scarf, and a pink shirt; a magazine.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair, wearing a pink scarf and a light pink shirt, sitting at a table and reading a magazine. These elements match the description provided. The presence of other items like scissors and papers on the table does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair, wearing a pink scarf, a pink pink shirt, sitting at a table and reading a magazine. The elements match the description provided. The setting of the elements in the and a on the table does not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..23ecbeafb62fbd519d72fee3c75f27ced47401cf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2be7a92e32d6df8a2a1d9aa7843dd73f7017446900a3e7c3495077c60529c25f
+size 88249
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..8ee111489eb3ca07b7dab4280455d8cf32f12cd1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of the engine compartment of a sports car. The engine is black and red, with a prominent red cover that stands out against the black components. The engine is well-maintained and clean, indicating that it is in good condition. The car's hood is open, revealing the engine and other components such as the air intake and battery. The background is blurred, but it appears to be an outdoor setting with a fence and grass visible. The style of the video is a straightforward, unembellished documentation of the car's engine, focusing on the details of the components and their arrangement. The video does not contain any text or additional graphics."
+    ],
+    "video_ids": [
+        "CWvEZm4chXw_22_31to221"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Engine (black and red), engine cover (red), air intake, battery? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a car engine with a black and red engine cover, a red engine cover, and visible air intake components. The battery is also present, located on the left side of the engine bay. All elements described in the condition are accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a car engine with a black and red engine cover, which red air cover, and an air intake components. There battery is not visible in as on the left side of the engine bay. The these match in the condition are visible depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cf3e0fe2124037457db7c8120674a870a22a595b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c10ec748e0513573f654c34531e8dea0d21e1effe476f768e379dcabb37666b3
+size 173822
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..d7ca93d9dbd41ac37f0aab2b81fd08e410bd6002
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a promotional advertisement for a silver Hyundai SUV. The car is shown in three different angles, showcasing its design and features. The first angle is a front view, highlighting the car's grille and headlights. The second angle is a side view, emphasizing the car's sleek lines and alloy wheels. The third angle is a rear view, focusing on the car's taillights and license plate. The car is parked on a tiled surface, and the background features a building and a tree. The video is designed to appeal to potential buyers, showcasing the car's design and features in a visually appealing way."
+    ],
+    "video_ids": [
+        "Evz5iUnf40Q_2_0to177"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A silver Hyundai SUV? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a silver Hyundai SUV, identifiable by its distinctive grille, logo, and overall design. The vehicle is the central focus, and all visual elements align with the description of a silver Hyundai SUV.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a silver Hyundai SUV from which by its distinct design, headlights, and overall design. The front is presented central focus of and the frames elements align with the description of a silver Hyundai SUV.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..81affae41f1f96cc9840f9cd51f78f0b8a9a02ca
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6320951c0976e30d25e6c3c01ac998fa127decf2ea5ecec428a918ab494a05f
+size 179960
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..3a4a3f7bbcc6e337484508905a2ad2911d23b8fe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red sports car with a black carbon fiber rear wing. The car is parked on a grassy area with a cloudy sky in the background. The car's design is sleek and modern, with gold rims on the wheels. The car's body is shiny and reflects the surrounding environment. The car's rear wing is a prominent feature, adding to its sporty appearance. The car is stationary, and there are no people or other objects in the immediate vicinity. The overall style of the video is a close-up shot of the car, focusing on its design and details."
+    ],
+    "video_ids": [
+        "UrmPokD8Fug_25_0to152"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red sports car with a black carbon fiber rear wing, gold rims on the wheels? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red sports car with a black carbon fiber rear wing and gold rims on the wheels, matching the description. The camera slowly zooms in on the rear of the car, highlighting these features without any conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a red sports car with a black carbon fiber rear wing and gold rims on the wheels, which the description provided The car angle zooms in on the car of the car, highlighting these features. any additional elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..492bf052e510f252dff5473204620f9299f1a0ac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79918768bdc2e195d53b088e2933f5f132e9cf39a064b8bbb18420905fcb27ce
+size 244883
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..855acb5213d68990bbf0bfdd65d4036826eb123d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a television interview featuring two men sitting in blue chairs on a stage with a red curtain backdrop. The man on the left is wearing a dark suit and white sneakers, while the man on the right is dressed in a black jacket and gray pants. Between them is a small round table with a globe on it. The man on the right is holding a microphone, suggesting he is the interviewer. The overall style of the video is formal and professional, typical of a television interview setting."
+    ],
+    "video_ids": [
+        "TreXqc4uDZk_2_110to244"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, a small round table with a globe, and a microphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men seated in chairs, facing each other, with a small round table between them that holds a globe. A microphone is also visible on the table near one of the men. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two men seated in chairs, a each other, with a small round table between them. has a globe. There microphone is also visible, the table, the of the men. The elements elements of in present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..829d8cc2818a72740083b4d3ee70a048246a33f0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:635741163d0c579dd18ca0409d69f42996ba4f9208888b7ae2bec260f21a1083
+size 72442
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..7f54e789f3c9d791c41df54f70b35c2d6d5ffeaa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a cartoon train character with a friendly face, driving past a wooden pole with a yellow plate hanging from it. The train is colorful with blue and yellow elements, and it has a red stripe around its middle. The train is on a track that runs parallel to the pole. The background is a simple, plain white, which puts the focus on the train and the pole. The style of the video is cartoonish and playful, aimed at a younger audience. The train's movement suggests it is in motion, and the yellow plate on the pole adds a touch of color to the scene. The overall impression is one of a fun, light-hearted animation."
+    ],
+    "video_ids": [
+        "TenfLILiolM_160_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A cartoon train character, a wooden pole with a yellow plate.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a cartoon train character with a smiling face, colorful design, and wheels, which matches the description. Additionally, there is a wooden pole structure with a yellow plate hanging from it, also matching the description. No elements contradict these core objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a cartoon train character, a blue face, which body, and a, which align the description of Additionally, there is a wooden pole with with a yellow plate attached from it, which matching the description. The additional contradict the core elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0cf8ea68189c348ba7ca546c8de610c3b13c6c22
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:42b6d8d4a7206ddeb27ff861b734cd203f02414a2fbc439b2c9fbf539e16dc67
+size 151713
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..df3a93acf0f136a54cd51ad8b072c8f269537d8e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a chef in a professional kitchen, preparing a meal. The chef is wearing a white chef's coat and is focused on the task at hand. The kitchen is well-equipped with various cooking utensils and ingredients. The chef is seen using a knife to chop vegetables, and a bowl is visible on the counter. The chef is also seen using a blender, suggesting that a smoothie or sauce is being prepared. The overall style of the video is realistic and informative, providing a glimpse into the daily workings of a professional kitchen."
+    ],
+    "video_ids": [
+        "XDvq231VB30_25_24to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chef, knife, vegetables, bowl, blender? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chef in a kitchen, mixing ingredients in a bowl. A blender is visible in the background, and there are vegetables (including colorful chopped ones on a tray) on the counter. All the specified objects are present, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person in a professional environment actively ingredients in a bowl. The knife is visible on the background, and there are vegetables onlikely broccoli bell items) a cutting) and the counter. The the elements objects are present and and the 'Object(s)' condition.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c9b2a4abeed80388314c764d4926cab40a7f7e9f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1218-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:31dd87b9d5e62bcbd9dd0723f5c1c54c101baea22dc2cb4b959c87145481fd08
+size 162845
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..b7108df1d4737b1a6ee63a9d28bb612740cecb5b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen cooking a dish in a wok on a stove. The dish appears to be a stir-fry, with the main ingredient being green leafy vegetables, possibly spinach or a similar type of greens. The person is using a wooden spoon to stir the ingredients in the wok, ensuring they are evenly cooked. The stove is visible in the background, with the burner beneath the wok glowing red, indicating that it is turned on and providing heat for the cooking process. The overall style of the video is a close-up shot of the cooking process, focusing on the wok and the ingredients being cooked, rather than the person cooking or the surrounding environment. The video captures the essence of home cooking, with the person actively engaged in preparing a meal."
+    ],
+    "video_ids": [
+        "YbcTZBvID8I_14_104to297"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a wok, green leafy vegetables (possibly spinach), a wooden spoon? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person adding green leafy vegetables (spinach) into a wok and stirring them with a wooden spoon. All the specified objects are present and their actions are consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a w cooking green leafy vegetables,possiblyach) into a wok. stirring them with a wooden spoon. The the elements objects are present and the actions align consistent with the description.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..571c96a8c2ca6bf21d62189a089c5268c248576c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a14fa161db668b8dbe74fd950b2d2fb0c15274d4a8ac5966f91e33a9840fd31a
+size 198999
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..3e2ce65eabb9a648ab1cc3705777e4f52bed5812
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a woman in a black wetsuit standing on a beach with a body of water in the background. She has long, dark hair that flows down her back. The wetsuit is tight-fitting and covers her entire body, including her arms and legs. The woman is holding a yellow tube in her hand, which is connected to a black hose that runs down her back. The hose appears to be part of the wetsuit, suggesting that it might be used for breathing underwater. The beach is sandy and the water is calm, with no waves visible. The sky is clear and blue, indicating good weather. The woman is the only person visible in the video. The style of the video is realistic and it appears to be a photograph rather than a painting or drawing. The focus is on the woman and her wetsuit, with the background serving as a simple, uncluttered backdrop. The lighting is natural and the colors are vibrant, suggesting that the photo was taken during the day."
+    ],
+    "video_ids": [
+        "WyTfmYzRONs_28_51to246"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Woman in a black wetsuit, long dark hair flowing down her back, holding a yellow tube connected to a black hose running down her back.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a black wetsuit with long dark hair flowing down her back. She is holding a yellow tube connected to a black hose that runs down her back, consistent with scuba diving gear. The core description is accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a black wetsuit with long dark hair flowing down her back. She is holding a yellow tube connected to a black hose that runs down her back. which with theuba diving equipment. The setting elements is accurately represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..eec9549a2e6b5b1b28293e744806b01a705c0f82
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1017241b45eb6a8462478f4f27898968bf8dce1218e6de466c731cf213e9542
+size 186700
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..2862d51de9e79bd2bf3ae592e99c8ba55d0e2ec3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with long hair, wearing a plaid shirt, standing against a white background. He appears to be speaking or reacting to something, as indicated by his open mouth and attentive gaze. The style of the video is a close-up shot, focusing on the man's face and upper body. The lighting is bright and even, highlighting the man's facial features and the texture of his shirt. The simplicity of the background puts the emphasis on the man and his expression. The overall impression is that of a casual, candid moment captured on camera."
+    ],
+    "video_ids": [
+        "6EW7hc2-Geg_17_0to144"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with long hair, wearing a plaid shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with long hair tied back in a ponytail and wearing a plaid shirt, which matches the description. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with long hair wearing back, a ponytail, wearing a plaid shirt. which matches the description provided The presence is lighting elements in not contradict the description description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7469937f1dac7e5dfc2d2e42305e3ab98e024b72
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:02aa65a6e0fbbd0fb7b3ee2efff1166a479fa7ae4a347461356e430e78306412
+size 323627
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..744cfa467dfe3bf881c887c5c61bf6b8202c44b0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a bald man wearing sunglasses and a scarf, sitting in a park. The man is the main subject of the video, and he is seen in various positions and expressions throughout the frames. The park setting is peaceful and serene, with palm trees and benches visible in the background. The lighting suggests it's daytime, and the overall atmosphere is calm and relaxed. The man's attire and sunglasses give him a stylish and sophisticated look. The video captures a moment of quiet reflection in a beautiful outdoor setting."
+    ],
+    "video_ids": [
+        "4ZeNq6s_aX4_0_0to116"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bald man wearing sunglasses and a scarf, sitting in various positions and expressions.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video consistently features a bald man wearing sunglasses and a scarf, seated in an outdoor park setting. His expressions and slight movements (like gesturing with his hand) vary slightly across frames, which aligns with the 'various positions and expressions' condition. The background elements, such as trees and benches, do not contradict the core description and are acceptable as additional context.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts shows a bald man wearing sunglasses and a scarf, sitting in a outdoor setting-like. The expressions and positions movements aligne turninguring with his hand) align,, the, which aligns with the descriptionsious positions and expressions' part. The background of, such as the and a, are not contradict the description description and are typical as additional elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..67c25c30e39832387083065dd0be7a18691af8df
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e76bfdb3bc12f682a246a01ec4aa79833072907f2b916addccca306b75e55889
+size 200642
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..9be225fece0d2c330777701c130bdb105510978f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a luxury car, showcasing its design and features. The car's interior is sleek and modern, with a focus on the dashboard and steering wheel. The dashboard is adorned with a large screen displaying a map, and the steering wheel is equipped with multiple buttons and controls. The car's interior is well-lit, highlighting the high-quality materials and finishes. The car is parked in a showroom, with a group of people gathered around it, admiring its design and features. The showroom is spacious and well-lit, with a high ceiling and large windows. The people in the showroom are dressed in business attire, suggesting that they are potential customers or industry professionals. The video is a promotional piece, designed to showcase the car's interior and features to potential buyers."
+    ],
+    "video_ids": [
+        "-SSfHBbNQV0_4_0to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Luxury car, dashboard, large screen displaying a map, steering wheel with multiple buttons and controls, people in business attire.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a luxury car with a modern interior. The dashboard features a large, curved screen displaying a map, and the steering wheel is visible with multiple buttons and controls. Reflections in the windows show people in business attire, confirming their presence around the car. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a luxury car interior a dashboard dashboard, The dashboard is a large screen central screen displaying a map, which the steering wheel is equipped with multiple buttons and controls. Thes on the windshield suggest a in business attire, indicating the presence in the car. The elements elements of in present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0ee6e2d87f5d99010b17a4ff36aee71b60bcbd49
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a7333571ad8897c66c65a70706b6c9cd602d8ce8c2692b94f4fd05dfe28aa036
+size 203207
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..539d78e2d30bf24c0b897a630b35f20b4ba49538
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a luxury sports car, showcasing its design and features. The car's interior is predominantly black, with leather seats and a dashboard that houses a variety of controls and displays. The car's door is open, revealing a yellow folder placed on the passenger seat, suggesting that the car is either being prepared for a journey or has just been used. The car is parked in a garage, with other vehicles visible in the background, indicating that the car is part of a collection or a showroom. The video is likely a promotional or sales video, aimed at highlighting the car's design and features to potential buyers."
+    ],
+    "video_ids": [
+        "wEzrR4vq3nU_51_0to128"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Leather seats, dashboard, controls and displays, a yellow folder on the passenger seat, other vehicles in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows the interior of a car with black leather seats, a dashboard with controls and displays, a yellow folder on the passenger seat, and other vehicles visible in the background through the windows. All elements described in the condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a interior of a car with leather leather seats, a dashboard with controls and displays, and yellow folder placed the passenger seat, and other vehicles in in the background. the car. The elements match in the condition are present in match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..771499081eb895f8de70d22c9eec477a16c83812
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7662549524b102a70db5fdbb087cbde0a3295e3c96856fd6b2f9cd63eb8f00dc
+size 182226
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..b3cc2488518f9b02439579746a5c98206f04d64f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men sitting at a table outdoors, engaged in a conversation. The man on the left is wearing a black sweater and has his hands clasped together on the table. The man on the right is wearing a blue plaid shirt and has his hands resting on the table. They are seated at a wooden table with chairs around it. In the background, there is a body of water with boats and a bridge, and the setting appears to be a riverside or lakeside cafe or restaurant. The atmosphere is casual and relaxed, with the men appearing to be enjoying their time together."
+    ],
+    "video_ids": [
+        "6jQ9wGMKDZk_4_0to162"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men sitting at a table, engaged in conversation. Their presence and interaction are the central focus, fulfilling the 'Object(s): Two men' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two men sitting at a table, engaging in conversation. The presence and interaction fulfill the central focus of fulfilling the 'Object(s)' Two men' condition. any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..778ffa167c29baff727c5fd23292732fd355ac18
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f920cd4242cc09da420b4a42246c0b1919518af674e26af574de094960c13537
+size 113560
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..535dfca99fc1dfe0faa8d6b7fbc4fa9d5bad25b6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a bowl of soup being enjoyed. The soup is served in a black bowl with a speckled pattern, and a silver spoon is placed within it. The soup itself is a rich orange color, suggesting a hearty and flavorful broth. The spoon is partially submerged in the soup, indicating that it is being used to enjoy the dish. The bowl is placed on a wooden table, which adds a rustic touch to the scene. The overall style of the video is simple and straightforward, focusing on the act of enjoying a warm and comforting meal."
+    ],
+    "video_ids": [
+        "iD4nLXnJhs0_64_59to195"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Black bowl with speckled pattern, silver spoon, soup (rich orange color).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a black bowl with a speckled pattern, a silver spoon resting inside, and a rich orange-colored soup with visible ingredients. These elements match the description exactly, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a black bowl with a speckled pattern, a silver spoon, inside it and a rich orange-colored soup. visible chunks like The elements match the description provided, fulfilling there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1d2ec98f31b419cad0d8f40d797a194f0d1bb9d5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1219-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2ec929945c8437606722025cf8656193d45c3022ee6a30da1b92bd49252908a6
+size 194425
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..49890000f099f02d7a857aeb4f29930a3fb467ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man swimming in the ocean. He is wearing a colorful hat with a red, green, and blue design, and he is wearing goggles. The man is swimming in a freestyle position, with his arms extended and his head submerged in the water. The ocean water is choppy, with small waves visible around the swimmer. The man appears to be swimming towards the shore, and the water around him is splashing as he moves. The video is likely taken during the day, as the lighting is bright and natural. The style of the video is realistic, capturing the man's actions and the surrounding environment in a clear and detailed manner."
+    ],
+    "video_ids": [
+        "CWD-nZ7rA5I_7_0to172"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man, colorful hat, goggles? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man swimming in the water, wearing a colorful hat and goggles, which matches the 'Object(s)' condition. Additional elements like a boat and water splashes do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man swimming in the water, wearing a colorful hat and goggles. which align the descriptionObject(s)' condition. The elements like the swim and seaw rashes are not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b6012fcf9778093224163b3862f7ff7995cdeef4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:31d1f731696c50c7319ec51118dc5ec499ae9b18da830570772b21acd787a40c
+size 263469
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..18323af0541e61498cf1b871f3480b70fc89b02d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a rectangular object placed on a wooden surface. The object appears to be a sharpening stone, characterized by its smooth, grayish-green surface and a distinct cut-out design resembling a stylized horse or similar figure. The stone is positioned diagonally across the frame, casting a shadow onto the wooden table beneath it. The lighting in the scene is soft, suggesting an indoor setting with natural light possibly coming from a nearby window. The wooden table has a warm, natural tone with visible grain patterns, adding texture to the background. There are no other significant objects or characters present in the video, and the focus remains solely on the stone and its surroundings. The camera remains stationary throughout the sequence, maintaining a consistent angle and framing that highlights the stone's details and the interplay of light and shadow."
+    ],
+    "video_ids": [
+        "ab018bae1d2c78ecfb9e874ae0407825fa18c0da6dbca64b845614ba87423db0"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A sharpening stone with a smooth, grayish-green surface and a cut-out design resembling a horse.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a rectangular sharpening stone with a smooth, grayish-green surface. There is a cut-out design on the stone that resembles a horse, as described. The stone is placed on a wooden surface, and no elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an rectangular objecting stone with a smooth, grayish-green surface. The is a cut-out design resembling the stone that resembles a horse, which described. The stone is placed on a wooden surface, which the additional contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..af16564fe6fd42e27944103de2489a0112499237
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f45ff64a0974705cc21280ea95450e0335c1c2379addab2f73a1dfd72808533
+size 44410
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..4003fc4811ecc105855a19023224b04cc9b7a253
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young man with glasses, wearing a blue denim shirt and a white baseball cap with a red and black design. He is seated in a room with a brick wall and a bicycle hanging on it. The man is gesturing with his right hand, which is blurred due to motion. His facial expression suggests he is in the middle of speaking or explaining something. The style of the video is casual and informal, with a focus on the man's actions and expressions. The room's interior suggests a personal or home setting."
+    ],
+    "video_ids": [
+        "7FfBzAMMZ8o_42_1018to1167"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man with glasses, wearing a blue denim shirt and a white baseball cap with a red and black design.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man wearing glasses, a blue denim shirt, and a white baseball cap with a red and black design, which matches the description. Additional elements like the brick wall and bicycle in the background do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man wearing a, a blue denim shirt, and a white baseball cap with a red and black design. which matches the description provided The elements like the background wall and the in the background do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..13c1f3dbf936e1e385c82ffe84b264d871936e8d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:44064e73423a0ed52e06544f13de195d4dff297ecd7f54534987ba33156500a2
+size 230840
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..bbc93b374bc699fdd579715c2fdf532cbc51282f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a brown bear is seen in a room filled with books and trophies. The bear is sitting at a desk, holding a piece of paper to its face as if reading it. The room has a vintage feel, with a brick wall and a poster of a circus hanging on it. The bear is the main focus of the video, and its actions suggest a sense of curiosity or interest in the paper it's holding. The room's decor and the bear's actions create a scene that is both intriguing and charming."
+    ],
+    "video_ids": [
+        "kqhZIFwTzz4_15_0to144"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A brown bear, a desk, a piece of paper, and a poster of a circus.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a brown bear sitting at a desk, holding and reading a piece of paper. Behind the bear, there is a poster with the word 'CIRCUS' and an image of a bear juggling, which matches the description. All core objects are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a brown bear sitting at a desk, holding and reading a piece of paper. The the bear, there is a poster of a image 'CIRCUS' visible some image of a circus,uggling, which align the description of The the elements are present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ba522e4e3b43eb969af98924dc9feac69333f4ce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aeea51cb283ec818c1248584e6db19c7913db8e7c45de66eaf41e46d8412f229
+size 159967
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..98226eb7b999cc63a77749c148e727423f6242c0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a journey on a winding road through a desert landscape. The road curves gently around the landscape, with a car driving on it. The car is a dark color, and it's moving away from the camera. The road is surrounded by red rock formations, which are scattered throughout the scene. The sky is clear and blue, and the sun is shining brightly. The overall style of the video is a combination of travel and nature, with a focus on the road and the surrounding landscape."
+    ],
+    "video_ids": [
+        "cAY0FG6zo4A_67_0to176"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A dark-colored car driving on the road.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dark-colored car (appears to be maroon or burgundy) driving along a curved road, which matches the core description. The surrounding scenery, including red rock formations and other vehicles in the distance, does not contradict this primary action.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dark-colored car drivingappears to be blackoon or darkundy) driving on a winding road in which align the description description of The road environment, including the rock formations and a vehicles, the distance, does not contradict the primary condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..34d26e3f87b0e07f9c4a204788daf24e4ea3bc06
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64030d9fa3c39ad2bc1abdc2264fb45d9996464d00cc508416c18dc891568baa
+size 241058
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..b7ddef81ef0bcd0bf11239a5f706a45386f05a34
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red car driving down a street. The car is a modern sedan with a sleek design, featuring a prominent grille and sharp headlights. The car is in motion, with the background slightly blurred due to the speed, indicating a sense of movement. The street is lined with buildings, and there are trees visible on the sides, suggesting an urban environment. The sky is clear and blue, suggesting a sunny day. The overall style of the video is dynamic and captures the car's motion effectively."
+    ],
+    "video_ids": [
+        "XNjjED4WerE_4_0to134"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red modern sedan? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red modern sedan driving on a road, which matches the description. The car's design is contemporary, and its color is clearly red. Background elements like buildings and trees do not contradict the core description of the object.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a red sedan sedan driving on a road. which matches the description of The car's design, contemporary, and the color is red red. The elements like the and trees are not contradict the core description of the object.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..99273ac3a97d87ebdcdc0e08b0679430285c5ef2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c46dabb723af8580cd9fb29e8147b9c57df9e7d4e591705c9111d9c81778239
+size 276002
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..e6eab8a75f825c8a71d4adf83b9fa79d682a6eab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a bird wading through shallow water. The bird, likely a sandpiper or similar shorebird, is seen from a side angle, showcasing its long legs and slender beak. Its plumage is predominantly white with black markings, which stand out against the dark, reflective surface of the water. The water's surface is rippled, with small waves creating a shimmering effect under the sunlight, indicating a calm yet dynamic environment. The bird moves gracefully, its legs partially submerged as it forages or simply explores the area. The overall atmosphere is tranquil, with the bird's movements being slow and deliberate, suggesting a peaceful moment in nature. There are no other significant objects or characters in the frame, focusing the viewer's attention solely on the bird and its interaction with the water. The lighting suggests it might be early morning or late afternoon, adding to the calmness of the scene."
+    ],
+    "video_ids": [
+        "56dde3db70d6f10c8899efde3bb231da9f9328afecc9d5b01c7d35eaa04d1ab8"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bird, likely a sandpiper, with white plumage and black markings.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bird that matches the description of a sandpiper with white plumage and black markings. The bird is seen wading in shallow water, which is consistent with sandpiper behavior. Its coloration and markings are visible, and while the lighting creates some reflections, the bird's features align with the given description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bird that appears the description of a sandpiper with white plumage and black markings. The bird is seen wading in water water, which is typical with thepipiper behavior. The appearanceation and the are clearly, and the the video and some reflections, it bird's features are with the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8824661b07788a11cdb9e784c6a02a70fc238dfb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6d6c38bfa3625cb3bcf5b55611777f3fa304cade9cb7e3401ddd41f88b9d129
+size 292781
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..055c7e620dc45a2ff886935fc5cc67d4aafd7b48
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a cartoon scene unfolds with a fox and two watermelons. The fox, with its red and white striped body, is standing on a rock, looking at the watermelons. The watermelons are green and striped, and one of them is being held by a character. The scene is set in a lush green forest, with trees and bushes in the background. The characters are animated, and the overall style of the video is cartoonish and colorful."
+    ],
+    "video_ids": [
+        "GbB01tc8lKg_51_75to266"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A fox and two watermelons.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a red fox standing on a rock, and two watermelons are visible \u2014 one large one being pushed by a character, and another smaller one held by the same character. The core objects described \u2014 a fox and two watermelons \u2014 are present in the scene, fulfilling the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a fox fox and in a forest, and there watermelons are present in one on and the held by the character in and another smaller one on by the fox character. The presence elements ( in a fox and two watermelons \u2014 are present and the video. fulfilling the '.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..722b538de095dc21facf2fd1e2c1fe91d34e3433
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-122-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4665277a71fb4af564703fae7536d0736ee85c915510b0ab435bf2c6b284c703
+size 185447
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..76e604d143ce7fd46d0c4a143e3c71a26637f15e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a culinary scene where three pieces of food, resembling sea creatures with spiky textures, are presented on a white rectangular plate. The food items are brown and have a glossy appearance, suggesting they might be a type of seafood or a seafood-inspired dish. In the background, there is a small bowl containing a dark liquid, possibly a sauce or a broth, which is placed on a wooden table. The style of the video is a close-up food presentation, focusing on the textures and colors of the food items, with a shallow depth of field that blurs the background slightly, emphasizing the food on the plate. The lighting is soft and warm, creating a cozy and inviting atmosphere. The video does not contain any text or additional elements that would provide context or information about the dish or the setting."
+    ],
+    "video_ids": [
+        "TdP4jItWRLA_25_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three pieces of food resembling sea creatures with spiky textures, a small bowl containing a dark liquid.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three pieces of food that resemble sea creatures with spiky textures, consistent with sea cucumbers, and a small bowl containing a dark liquid, likely a dipping sauce. These elements match the description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three objects of food that have sea creatures with spiky textures, placed with the umbers or which a small bowl containing a dark liquid, which representing sauce sauce or The elements match the description provided any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4f69ab5c98c8c3026d5433061d5380bc7aca65ae
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae5801851759058c6f0d6fbc879bf928367556c0477e750acfc6c47aa263bbad
+size 42831
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..7c4edca327d5bcc3d6f27b3fc4275305ac61d4be
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car, focusing on the driver's side. The car has a sleek design with a black and red color scheme. The steering wheel is black with a red center, and the dashboard is also black with red accents. The car's interior features red leather seats and a black center console. The car's door is open, revealing a red interior door panel. The car's door handle is black, and the window controls are located on the door panel. The car's door lock is also black. The car's door is closed, and the window is rolled up. The car's door is open, and the window is rolled down. The car's door is closed, and the window is rolled up. The car's door is open, and the window is rolled down. The car's door is closed, and the window is rolled up. The car's door is open, and the window is rolled down. The car's door is closed, and the window is rolled up. The car's door is open, and the window is rolled down. The car's door is closed, and the window is rolled up. The car's door is open, and the window is rolled down. The car's door is closed, and the window is rolled up. The car's door is open, and the window is rolled down. The car's door is closed,"
+    ],
+    "video_ids": [
+        "Tsq25yakIzM_10_0to160"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ['Steering wheel', 'Dashboard', 'Red leather seats', 'Black center console', 'Red interior door panel', 'Door handle', 'Window controls', 'Car door lock']? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully displays all the specified objects: the steering wheel is prominently visible, the dashboard is shown with its digital display, red leather seats are clearly depicted, the black center console is present, the red interior door panel is visible, and the door handle and window controls are discernible. The car door lock is not explicitly shown but is not a critical omission for fulfilling the core description. All elements align with the requested objects without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful the the objects objects: the steering wheel, visible visible, the dashboard is shown in a controls display, the leather seats are clearly seen, the black center console is in, the red interior door panel is visible, the the door handle and window controls are alsoible. The car door lock is also explicitly shown, is implied necessary requirement part as the the ' description.\"\n The elements are with the provided objects.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..eea7d9c0f9043e05e8dbaba6a468c29acefa9f7c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d75be46a98c9c0d95187ac4c6ad108e8092f58d43c2217142819e40b5c81943e
+size 204347
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..04c06f44baa321d85162e8feae410087cace674f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up of a small kitten with striking blue eyes and a mix of gray, black, and orange fur. The kitten is lying on a blue surface, which could be a blanket or a cushion. The kitten's eyes are wide open, and it appears to be looking directly at the camera, giving the impression of curiosity and innocence. The kitten's fur is soft and fluffy, and its ears are perked up, indicating alertness. The kitten's tail is curled up, adding to the overall cuteness of the scene. The video is likely to be a short, sweet, and adorable clip that would appeal to cat lovers and anyone who enjoys watching cute animals."
+    ],
+    "video_ids": [
+        "gdZs3ciRq0s_42_0to139"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small kitten with blue eyes and mixed fur (gray, black, and orange).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small kitten with large, prominent blue eyes and a coat that clearly features a mix of gray, black, and orange fur, matching the description. The kitten is also positioned on a blue cushion, which does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a small kitten with blue, expressive blue eyes and fur fur that displays displays a mix of gray, black, and orange fur. which the description provided The kitten's lying lying in a blue surface, which is not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f777a949f9e8ff2a3440640645c06528f086ca5f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a48e1561600c5779a414e8cf1850717a158bcb960937488094a46e6284f0282b
+size 212797
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..83d651e0ef3ca9f55b82da9dd9d5343656303e87
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a sequence of events in a car showroom. In the first frame, a sleek black sports car is parked prominently in the foreground, its shiny exterior reflecting the showroom's lights. The car's design is modern and aerodynamic, with a low profile and large wheels.  In the second frame, the camera angle shifts slightly, revealing a red sports car parked next to the black one. The red car is smaller in size but equally sleek, with a vibrant color that stands out against the black car.  In the third frame, the camera angle changes again, showing a third sports car parked in the background. This car is a silver color, with a more traditional design compared to the other two. It's parked further back in the showroom, giving the viewer a sense of depth and space.  Throughout the video, the showroom itself is visible, with its high ceilings and large windows allowing natural light to flood in. The cars are the main focus, but the showroom's design and lighting also play a role in creating an atmosphere of luxury and sophistication."
+    ],
+    "video_ids": [
+        "toAfCELykCU_34_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ['sleek black sports car', 'red sports car', 'silver sports car']? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a sleek black sports car in the foreground, a red sports car partially visible in the lower left, and a silver sports car in the background. All three specified cars are clearly present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a sleek black sports car in the foreground, which red sports car in visible in the background part corner and a silver sports car in the background. The three cars types are present present, match the description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6f017aefbc2a2852df8e8a84b9ab24ddfcff85c3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc6d3ccb42bad235da2fd883cef7464708f8b39e407c6d196fd10d02acd3bc99
+size 67768
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..e289a0eee741862dc466b5568ba578f8e0627214
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a monkey wearing a red jacket and a black top hat is riding a brown horse. The monkey is also wearing glasses. The horse is walking on a dirt road. In the background, there are trees and hills. The monkey is leading a line of white rabbits, which are following the horse. The rabbits are wearing red jackets and black top hats, just like the monkey. The scene is set in a colorful and cartoonish style."
+    ],
+    "video_ids": [
+        "DzvkD8qyNFA_163_0to117"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A monkey, a brown horse, and a line of white rabbits.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a monkey wearing a black hat and glasses, riding a brown horse. Behind them, there is a line of white rabbits dressed in red uniforms, marching in formation. These core elements match the description exactly, and no contradictory elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a monkey riding a red top and a, riding a brown horse. There the, there is a line of white rabbits. in clothes coats. which in a. The elements elements match the description provided, and there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6aaa54d2f5bee65f746801f75e3fb0d1b0e2bdc9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:317d4e3aee84edefa0df8ec91b0bff66887e7cc4fb1e960d4bc9b5484ae2c75c
+size 119365
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..8013af7423a23f872213a7ddd9ea2d51e22c8724
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a close-up view of a red and silver Cosmis XT wheel, which is the main focus of the image. The wheel is positioned in the center of the frame, with its shiny silver spokes and red center drawing attention. The background is filled with other wheels, creating a sense of depth and perspective. The wheels in the background are slightly blurred, emphasizing the Cosmis XT wheel in the foreground. The overall style of the video is sleek and modern, with a focus on the wheel's design and details. The lighting is bright, highlighting the wheel's metallic finish and the contrast between the red and silver colors. The video does not contain any text or additional objects, keeping the viewer's attention solely on the Cosmis XT wheel."
+    ],
+    "video_ids": [
+        "M_O76buJKMU_18_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red and silver Cosmis XT wheel (main focus)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red and silver Cosmis XT wheel as the main focus, with clear visibility of its design, branding, and details. The background elements, such as other wheels, do not contradict the core description and are consistent with a setting like a wheel showroom or workshop.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red and silver Cosmis XT wheel, the main focus. with a details of the design and including, and the. The background,, while as other wheels, do not detr the core description and serve secondary with the display that a wheel display or display.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f08549a3b201b8ea278da6c1333858ae138b4cf2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:92cd956deb94a0c400752c6c76703cf1d2fa3ad8e8a0058ede83063be9681494
+size 86788
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..3fe3dfb729e48b1c79db3c87054f94d26c560fcd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up shot of a bird in flight, showcasing its intricate details and dynamic movement. The bird, likely a pigeon or dove, is seen from below, emphasizing its wings and tail feathers. Its plumage displays a mix of white, gray, and brown hues, with the wings spread wide, revealing the texture and pattern of the feathers. The bird's head is turned slightly to the side, giving a glimpse of its beak and eye, which appear alert and focused. As the bird soars through the air, its wings flap rhythmically, creating a sense of motion and grace. The background is blurred, suggesting a natural outdoor setting with soft, diffused lighting that enhances the bird's features. The BBC logo in the top left corner indicates the source of the footage, while the number \"8\" in the top right corner suggests it might be part of a series or episode. The overall scene conveys a serene and majestic moment in nature, highlighting the beauty and agility"
+    ],
+    "video_ids": [
+        "fe400e488bcfb8a746c7847813ca125401d1fa90fa393187d41914a9921c9fe1"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bird (likely a pigeon or dove)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bird in flight with the characteristic features of a pigeon or dove, including its body shape, wing structure, and coloration. The bird is clearly visible and matches the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a bird in flight, features characteristics features of a pigeon or dove, such its shape shape, wings structure, and plumation. The bird's depicted the and matches the description of.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7f36986b2f9d7e54413a3b05d26bfb30c7cf3e1a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3131ed6cc65d3917885b5cd7947856794e72c12630eee490f11e1eb0a68d228d
+size 227285
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..f88a100b6527fce616888604f56577825deee94f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a vibrant city scene with a police station as the focal point. A yellow helicopter is seen landing on the roof of the police station, indicating a swift response to an emergency. The police station is surrounded by a variety of buildings, including a brick building and a white building, showcasing the diversity of architecture in the city. A truck is parked in front of the police station, possibly for maintenance or support purposes. The cityscape is further enhanced by the presence of mountains in the background, adding a touch of nature to the urban setting. The overall style of the video is realistic with a slight touch of animation, giving it a unique and engaging look."
+    ],
+    "video_ids": [
+        "5sCZJN0ITt4_14_0to104"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Yellow helicopter, police station, brick building, white building, truck? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a yellow helicopter hovering above a police station, which is marked with the word 'Police'. There is also a brick building adjacent to the police station and a white building that appears to be part of the station complex. A white truck is visible parked in front of the police station. All specified objects are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a yellow helicopter, above a building station, which is a by a word 'POL' There are a a brick building and to the police station and a white building nearby appears to be a of the police complex. A truck truck is visible in in front of the police station. The the objects are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..077e14bc1ced26c3eb9627d547d770e31edf7805
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1220-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fcfa50280ec678e262a164574ac1cd14838ae1ceb539a2addcfc60bf495fe793
+size 231585
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..b2ee74d265efd71e2ac4d61e442569527d8cc49e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the grandeur of a multi-level shopping mall. The camera pans from the top floor, showcasing the intricate design of the ceiling, adorned with a colorful stained glass mosaic. The mall's architecture is highlighted by the white pillars and arches that support the structure. The camera then descends to the lower levels, revealing the bustling activity of shoppers and the vibrant display of storefronts. The video encapsulates the lively atmosphere of a modern shopping center, with its blend of architectural elegance and commercial vibrancy."
+    ],
+    "video_ids": [
+        "rUBBuRG9Cl4_98_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Colorful stained glass mosaic, white pillars, arches, shoppers, storefronts? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a colorful stained glass mosaic on the ceiling, white pillars supporting the structure, arched walkways, shoppers moving through the mall, and storefronts visible on various levels. All described elements are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as showcasing showing a colorful stained glass mosaic in the ceiling, white pillars, the structure, andched windowsways, and moving through the space, and storefronts visible in the levels. The these elements are present and contribute with the video content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0e28fb286d4c08d4f9668f8182c57c107bb0ae95
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c60defe2c972759a9f329666e05da6dad0454541a0bea869f6d3a2a90dce8119
+size 440068
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..08c6938f46d36597ba80991473315969b0a9b79f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a family is gathered around a dining table in a cozy kitchen. The father, wearing a cowboy hat, is seated at the table with his daughter, who is dressed in a bear costume. The mother stands nearby, holding a plate of food. A blue cat-like creature stands on its hind legs, reaching out towards the plate of food. The creature's fur is a vibrant blue, and it has a friendly expression on its face. The kitchen is filled with warm, inviting colors, and a clock hangs on the wall, indicating the time. The family seems to be enjoying a meal together, with the blue creature eagerly participating in the feast."
+    ],
+    "video_ids": [
+        "1eV2Vwgvax8_131_0to143"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A family (father, daughter in a bear costume, mother), a blue cat-like creature? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a family consisting of a father (wearing a cowboy hat and red plaid shirt), a daughter in a bear costume (orange, sitting on a chair), and a mother (wearing glasses and a pink shirt). Additionally, there is a blue cat-like creature (a small blue character with cat ears) present, which matches the description. No elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a family consisting of a father,wearing a cowboy hat), a shirtaid shirt), a daughter in a bear costume,w and white at the high), and a mother (wearing a and a blue top) Additionally, there is a blue cat-like creature sittinga plush stuffed stuffed) a-like and present at which align the description. The elements contradict the setup description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..26611dce823cd3bde132de1eacd57a2554f2444b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a936c13d99cf104a238e4b9bbfb2f1506716b7c60ff3b11d3eb22e9d2c0bf5b4
+size 153864
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..8e28ba9e79a01ec394ec8c534223df6110d0f929
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a lively scene with a green parrot and a gray cat. The parrot, with its vibrant green feathers and yellow beak, is the main focus of the video. It is holding a red guitar and appears to be playing it, adding a sense of music and fun to the scene. The cat, with its gray fur and pink nose, is standing next to the parrot, seemingly enjoying the music. The background is filled with yellow confetti, creating a festive atmosphere. The video is likely a 3D animation, given the detailed and colorful characters and the dynamic background. The overall style of the video is playful and entertaining, with a focus on the interaction between the parrot and the cat."
+    ],
+    "video_ids": [
+        "xnkOs7Hi_eY_5_54to268"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A green parrot and a gray cat? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a green parrot holding a guitar and a gray cat with green eyes, both of which match the described objects. The presence of confetti and a red heart balloon does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a green parrot and a red and a gray cat sitting a eyes standing which of which match the description objects. The presence of aetti in a colorful guitar in does not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b24a7efd30b0927370b082fde5b22600235868b0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2d016565c38d72148f7aa6c95ede9d5327e589aadb06475aaa7de76d88bbf2dd
+size 336354
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..35d8c3605f0a9c19f906db83208c4ae0bd8bf6b0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two Jeep SUVs on a dirt road in a mountainous area. The first Jeep is red and is parked on the left side of the road. The second Jeep is yellow and is parked on the right side of the road. Both Jeeps have their headlights on. The road is unpaved and appears to be in a rural or wilderness area. The background features a mountainous landscape with trees and rocks. The style of the video is realistic and it captures the natural beauty of the area."
+    ],
+    "video_ids": [
+        "QlppU7Re5_M_9_0to119"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two Jeep SUVs? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two Jeep SUVs, one red and one yellow, parked on a dirt path in a mountainous area. The vehicles are the central focus and match the description of 'Two Jeep SUVs' without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two Jeep SUVs, one red and one yellow, parked on a beach road. a mountainous area. The presence are the primary focus of fulfill the description of theTwo Jeep SUVs'. without any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e134017e3c066fa0151a671e4004053c7bc3de06
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:637bbc20efec43ca4b24349d189b09ff543ab71dd17b7ee04294e7622335e62b
+size 265589
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..6702cdd3527ceecf11d25debc3be62c19066fe9b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man in a plaid shirt and overalls, wearing a cowboy hat, interacts with two children in a rural setting. The man is holding a red cup and appears to be pointing at something. The children, a boy and a girl, are standing next to a blue bucket. They are looking at the man with interest. The background features a wooden bench and a chair, suggesting a relaxed, outdoor setting. The overall style of the video is cartoonish and colorful, with a focus on the characters and their interaction."
+    ],
+    "video_ids": [
+        "2oopbQOW7WM_334_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a plaid shirt and overalls with a cowboy hat, a boy, a girl, a red cup, and a blue bucket.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a plaid shirt and overalls with a cowboy hat, a boy, and a girl. A blue bucket is visible on the ground, and the man is holding a red can labeled 'BIG FAT WORMS', which can be considered a red cup in context. All specified objects are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man wearing a plaid shirt, overalls with a cowboy hat, holding girl, and a girl. The red bucket and visible, the ground, and the man is holding a red cup. 'cup'. SASTES'. which is be interpreted a red cup. this. The elements elements are present in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b59f9389aed480c6ccf049cf9919a32a8234d11f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f18c796a3c892dd2b3d045202d21c694cc7ce532647224c42a07a965c5024be
+size 166983
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..9b61faf382f80adff0bf36d81278da9d00a7f02f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a black cast iron pot placed over an open campfire. Inside the pot, a vibrant dish is being cooked, featuring chunks of meat, possibly chicken or pork, mixed with vegetables such as onions and possibly bell peppers. The dish appears to be seasoned with a yellowish-orange spice, likely turmeric or curry powder, giving it a rich and appetizing color. Steam rises from the pot, indicating that the food is hot and actively cooking. The pot is suspended over the fire using a metal stand, which is supported by a piece of wood. The background shows a natural outdoor setting with scattered logs and some greenery, suggesting a camping or outdoor cooking scenario. The camera remains stationary throughout the video, focusing on the pot and its contents, allowing viewers to observe the cooking process in detail. There are no visible characters or animals in the frame, and the primary focus remains on the food and the cooking environment."
+    ],
+    "video_ids": [
+        "ef14138231e878dae72fb447afcb0d575bc4bdebe5ab68fcd2b2a05dd61d7712"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black cast iron pot, a vibrant dish inside the pot, a metal stand, and a piece of wood supporting the stand.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a black cast iron pot suspended over a fire, containing a vibrant dish with visible ingredients like meat and vegetables. A metal stand is visible supporting the pot, and a piece of wood is used to hold the stand, consistent with the description. The scene is set outdoors with logs and smoke, which aligns with the context of outdoor cooking.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a black cast iron pot with over a fire, containing a vibrant dish with various ingredients such corn and vegetables. The metal stand is supporting, the pot, and a piece of wood is supporting to support the stand in which with the description. The presence is outdoors outdoors, a and green, enhancing addss with the camping of a cooking.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2bc22e93c223603714ab8817b7577535d9718cea
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:94b4b69171f133057a7b992063ff13fb289cdfef6e2524e5efc7a70ea85026ab
+size 126040
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..8e9603ea910f744cf7bd991662ff11549a9ce334
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a rusted, burnt-out car in a junkyard. The car is in a state of disrepair, with its body and interior severely damaged by fire. The car is surrounded by other junked vehicles and debris. The junkyard appears to be located in an industrial area, with a large truck visible in the background. The overall style of the video is gritty and realistic, capturing the harsh reality of the junkyard environment."
+    ],
+    "video_ids": [
+        "H881j5wmIbo_20_1347to1485"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A rusted, burnt-out car, other junked vehicles, and debris.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a rusted, burnt-out vehicle that appears to be a car or truck frame, with visible signs of fire damage and rust. Surrounding it are other junked vehicles and debris, including parts of trailers and other metal scraps. The scene matches the description of a junkyard or scrapyard setting with the specified objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a collectioned, burnt-out car in appears to be a car, truck,, surrounded other signs of decay damage and rust. Surrounding this are other vehiclesed vehicles, debris, which what of cars and other damaged scraps, The scene is the description of a junkyard with ayard with, the presence objects.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b406593eee7374369db1ed00f90c9804b43c253c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a16ea05d2ca6350b10e8c71ceb69b3c1c647eb1f8a307e234abd2693b8034e5
+size 241187
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..8b1293e59549f122503ae58e2b00f2888114122a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a casual, candid snapshot of a man lying in bed, seemingly in the middle of a video call. He is shirtless, with a relaxed posture, and appears to be engaged in a conversation with someone on the other end of the call. The person on the other end of the call is a woman, who is also visible in the video. The style of the video is informal and personal, capturing a moment of everyday life. The focus is on the interaction between the two individuals, with the setting serving as a backdrop to their conversation."
+    ],
+    "video_ids": [
+        "4L9ibDDrtkA_1_309to518"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man (shirtless) lying in bed, Woman (on the video call).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a shirtless man lying in bed, and there is a video call window displaying a woman, identified as Emma Chamberlain. These elements match the 'Object(s)' condition described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a shirtless man lying in bed, and a is a woman call with displaying a woman. which as the.lain. The elements match the descriptionObject(s)' condition provided in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f1ee38ad9195c14504db54a52916d6eedcd8504f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1221-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae377bcbc3250c25e4d1b7d290e03debd96b9d359850625c6a8e3a4e2a7fc4a9
+size 148463
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..aa00fb5a5d1f6f81f38f57a44a0524c074fdf4d3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a black and yellow motorcycle parked on a sidewalk next to a tree. The motorcycle is parked in front of a black metal fence. The tree has a blue rope tied around it. The motorcycle is parked on a brick sidewalk. The motorcycle is parked in front of a white building. The motorcycle is parked on a sidewalk next to a tree. The motorcycle is parked on a sidewalk next to a tree. The motorcycle is parked on a sidewalk next to a tree. The motorcycle is parked on a sidewalk next to a tree. The motorcycle is parked on a sidewalk next to a tree. The motorcycle is parked on a sidewalk next to a tree. The motorcycle is parked on a sidewalk next to a tree. The motorcycle is parked on a sidewalk next to a tree. The motorcycle is parked on a sidewalk next to a tree. The motorcycle is parked on a sidewalk next to a tree. The motorcycle is parked on a sidewalk next to a tree. The motorcycle is parked on a sidewalk next to a tree. The motorcycle is parked on a sidewalk next to a tree. The motorcycle is parked on a sidewalk next to a tree. The motorcycle is parked on a sidewalk next to a tree. The motorcycle is parked on a sidewalk next to a tree. The motorcycle is parked on a sidewalk next to a tree. The motorcycle is parked on a sidewalk next to a tree. The motorcycle is parked on a sidewalk next to a tree. The motorcycle is parked on a sidewalk next to a tree."
+    ],
+    "video_ids": [
+        "hFm8cyxwN8U_54_0to128"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black and yellow motorcycle? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a motorcycle that is predominantly black with distinctive yellow front forks, matching the 'black and yellow motorcycle' description. The other elements in the scene, such as trees and a fence, do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a motorcycle that is predominantly black with yellow yellow accents f and which the descriptionblack and yellow'' description. The presence elements in the video, such as the, a building, do not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d773b8eaf33d9515306b176121ecfb51a11274e6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40ea1ecccd4b26e3207b166f913a5db9472a8d7f39378ff001686cdabc0c3079
+size 127742
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..e0dc026a5274f26c996c135bdfbfab8510539bd4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a lively street scene at night. A woman, dressed in a green jacket, is seen holding a microphone, presumably conducting an interview. She is standing in front of a man who is wearing a white shirt with a green logo. The man appears to be the subject of the interview. The background is bustling with people, adding to the vibrant atmosphere of the scene. The video seems to be a news report or a street interview, capturing a moment in the life of the city."
+    ],
+    "video_ids": [
+        "b2rCuQVAT1M_5_376to513"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a green jacket holding a microphone and a man wearing a white shirt with a green logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a green jacket holding a microphone and a man wearing a white shirt with a green logo (the letter 'P' on his shirt). These elements match the description provided. The background crowd and setting do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a green jacket holding a microphone and a man wearing a white shirt with a green logo.the letters 'ING' is the shirt). The elements match the description provided, The presence and and nighttime do not contradict the core description but}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e8ae7b9cbd7a8c31c6572fbf6fc27ea5f34e9d05
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5d805dec8892cf67a66ad2634d15ce736539812236f4b96b378584b79854db6
+size 192336
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..a23f75e72fe8bd0f9e67a066a02442600b9a0006
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man sitting at a desk with a pair of black headphones on a stand in front of him. He is wearing a black hoodie and appears to be speaking or gesturing with his hands. The desk has a laptop and a few other items on it. In the background, there are potted plants and a window. The style of the video is casual and seems to be a personal or informal setting."
+    ],
+    "video_ids": [
+        "DmsRGuK7P_Y_13_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a pair of black headphones on a stand, a black hoodie, a laptop, and other items on the desk.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man sitting at a desk, wearing a black hoodie. On the desk, there is a pair of black headphones on a stand, a laptop, and other items such as a black case and a charging dock. All core objects mentioned in the condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing at a desk with wearing a black hoodie, In the desk, there is a pair of black headphones on a stand, a black, and some items that as a p object and a plant cable. The these elements mentioned in the description are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..97be9c9ef44cb3f84af0ead6403f05ae3eadebff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e20048e15bd10b69c8eefd169c4b15e5192f3a86486ae8ef527618607a85928
+size 120744
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..f7870f7332fca0cc67fe0bc92d4a51e3f36f2d30
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young girl with blonde hair and a pink hat is seen interacting with a large, furry hand. The hand is brown and has long, sharp claws. The girl is wearing a pink and purple striped dress. The background features a plaid curtain. The interaction between the girl and the hand appears friendly and curious. The video is likely a scene from an animated film or show, with a focus on the interaction between the girl and the creature with the large hand."
+    ],
+    "video_ids": [
+        "Hz5sjIMafDc_93_42to199"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young girl, a large, furry hand? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a young girl wearing a purple beret and a striped outfit, interacting with a large, furry hand. The hand belongs to a bear, which is also visible later in the video, confirming the presence of the described objects without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a young girl interacting a pink dresset and a pink dress, interacting with a large, furry hand. The hand is to a character, which is a AI in in the video. fulfilling the presence of a ' object. any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..53ebbbac67e34f38d252b9e3728b0fd30ada9115
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d1a3473d9ccc8a735b60de8c6f65b4f45046461f9f6ccd7efb76f0832f9c4ebf
+size 234116
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..80bd8c421fb6a145145176d52c44085db680a3f2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a teal-colored Volkswagen electric car parked on a city street. The car is sleek and modern, with a distinctive design that includes a large front grille and angular headlights. The car is parked next to a building with a dark facade, and the street appears to be quiet and empty. The car is the main focus of the video, and it is shown from a side angle, allowing viewers to see its design and features. The overall style of the video is clean and minimalist, with a focus on the car and its surroundings. The lighting is soft and natural, suggesting that the video was taken during the day. The car's color stands out against the darker background, drawing attention to its design and features."
+    ],
+    "video_ids": [
+        "2oHywmuUk44_18_0to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Teal-colored Volkswagen electric car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a teal-colored Volkswagen electric car, specifically the 2020 Volkswagen I.D. 3, which matches the description. The car is shown in clear detail, and its color and model are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a teal-colored car electric car. which a ID2022 Volkswagen ID.D. V3 model which is the description. The car is shown in a detail, including the electric and model are consistent represented.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f9daff8142e336f702409eb4e791e6e3fcfced32
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cfd256727eede2ac67288f6928c95a8fa8043f492b83eaa76847ebcabe64bfc5
+size 167830
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..c7b85a36211f8c80895b9edfac97b2cf67ae7e9f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene on a large body of water, possibly an ocean or a large lake. The sky is filled with clouds, and a rainbow arcs across it, adding a touch of color to the otherwise blue and white sky. Several sailboats are scattered across the water, their sails catching the wind and propelling them forward. The boats vary in size and design, suggesting a diverse group of sailors. The water is calm, with only small ripples disturbing its surface. The overall style of the video is peaceful and tranquil, capturing the beauty of nature and the joy of sailing."
+    ],
+    "video_ids": [
+        "bytBdlYT0ZU_4_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Several sailboats? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows multiple sailboats on the water, including a prominent catamaran in the center-right, another sailboat on the left, and additional sailboats visible in the distance. This matches the 'Several sailboats' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows several sailboats on the water, which a central oneamaran in the foreground and and which catboat to the left, and a sailboats in in the background. The align the descriptionSeveral sailboats' condition as}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1bbecc7b20c6bc2c69b22f6cfcce7cafafb141de
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d96349b126e2d86374a99c8530f436c641e064e292d222604050bd9d00f8b54
+size 161186
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..e5e90f7a423691d6490a44b826573334e8c3426a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a blue Cobra Jet car on display at a car show. The car is adorned with a striking snake design on its side, adding a touch of uniqueness to its appearance. The car is parked in a spacious showroom, surrounded by other cars, indicating a variety of vehicles on display. The car's sleek design and vibrant color make it stand out amidst the other vehicles. The showroom's lighting highlights the car's features, drawing attention to its details. The video captures the essence of a car show, where enthusiasts gather to appreciate the craftsmanship and design of various vehicles."
+    ],
+    "video_ids": [
+        "LygKiHO2eiE_5_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue Cobra Jet car with a striking snake design.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a blue Cobra Jet car with a striking snake design (a cobra emblem) on its side, matching the description. Additional cars and exhibition elements in the background do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a blue car Jet car with a striking snake design ona stylobra)) on the hood. which the description provided The elements and a-like in the background do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..359e2fc4b4c472378fb73fc7a42d2532c0c5b0d1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aafb8e66085a1beccee8047a95811c9ae32dce47a49779c6a66778e73e80f6dd
+size 223658
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..7e22c9dc00176db0ea984c8030f2d5d507deb251
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a small, cozy cabin with a wooden interior. The cabin features a desk with a lamp and various electronic devices, including a radio and a phone. A bench is situated next to the desk, providing a place to sit. The cabin appears to be well-equipped for communication and work, suggesting it might be used for remote work or as a command center. The overall style of the video is simple and functional, focusing on the practical aspects of the cabin rather than aesthetic embellishments."
+    ],
+    "video_ids": [
+        "60iSUzmmyBc_2_102to224"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A desk with a lamp and various electronic devices (radio, phone), a bench? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a wooden desk with a lamp attached to its side, and above the desk, there are various electronic devices including a radio and what appears to be a phone or communication device. A bench with a cushioned seat is also visible next to the desk, matching the described elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a desk desk with a lamp and to it side. which various the desk, there are various electronic devices, what radio and what appears to be a phone. similar device. There bench is a cushion is seat is also visible beneath to the desk. fulfilling the description elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e0bedeba0ff7a1b2b9183ad95ddd13ea08f3fffb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1222-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59a695b5e7150fc3f0c1cada9dca7e0493957cf9f7766a67c2c16812a8639382
+size 126313
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..b58271398b0b56c8e9e1ad6f46be249e6a14c449
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the rear view of a sleek, black sports car in motion. The car's design is modern and aerodynamic, with a distinctive rear light that features a unique, angular design. The car's body is shiny and well-maintained, reflecting the light as it moves. The car is driving on a road, and the background is blurred, indicating the car's speed. The overall style of the video is dynamic and stylish, emphasizing the car's design and movement."
+    ],
+    "video_ids": [
+        "KS0vG2GNCzY_24_0to103"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A sleek, black sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a sleek, black sports car, specifically a Lamborghini, with distinctive angular design and red taillights, matching the description. The car's matte black finish and sporty aesthetic are clearly visible, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a sleek, black sports car in which focusing rearorghini Hur as a features ta elements a taillights. which the description of The car is design black finish and sporty features are consistent visible, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c762d79a901fae8440ca3de6a3486a316a99ab96
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57012ee72f601d13d2e50a036cc9f37908f5221af4b14f562db54dae61034d85
+size 225912
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..5734958007ef49c963dc860d5ee39a7ec83abd0f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with long dark hair and a gray shirt. She is standing in front of a red car and a white tent. In the first frame, she is looking directly at the camera with a surprised expression. In the second frame, she is still looking at the camera but her mouth is open as if she is about to speak. In the third frame, she is gesturing with her hands, possibly explaining something or emphasizing a point. The style of the video is casual and candid, capturing a moment in the woman's day."
+    ],
+    "video_ids": [
+        "-ibOcP8szbg_16_114to293"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with long dark hair and a gray shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman with long dark hair and a gray shirt, which matches the description. The background elements, such as cars and an air conditioner, do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a woman with long dark hair wearing a gray shirt. which matches the description provided The background includes, such as the and green outdoor conditioning, do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2e021854c7d1c539e21eefc7d40326c667a5a755
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:774312a66371b0282a1874f9597c41f780591cdcf8708f48d5a32827765406dd
+size 146389
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..3720fd130441d3496236b012cf32c90bec55629d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene outdoor scene featuring two birds interacting with bird feeders on a sunny day. The background is a lush green, indicating a garden or park setting. In the foreground, there is a metal bird feeder with a conical roof, filled with a variety of seeds and nuts. To the left, an orange clay bowl contains a mix of seeds and grains. A small bird, likely a sparrow, is perched on the edge of the clay bowl, pecking at the food. Another bird, also appearing to be a sparrow, flies into the frame from the right side, landing on the clay bowl and joining the first bird in feeding. The second bird flaps its wings briefly before settling down to eat. The camera remains stationary throughout the sequence, focusing on the interaction between the birds and their food sources. The lighting is bright, suggesting it is daytime, and the overall atmosphere is calm and natural."
+    ],
+    "video_ids": [
+        "b62ba86563fdf9f6f0a2c4274d172571ab962430cbeca19c6cc4c9c159fad2e3"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two sparrows, a metal bird feeder with a conical roof, and an orange clay bowl.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two sparrows, one on the left feeding from an orange clay bowl and another landing nearby. A metal bird feeder with a conical roof is centrally positioned, filled with nuts. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two birdsrows interacting a of a metal and from a orange clay bowl and the on on. There metal bird feeder with a conical roof is visible located in and with seeds. The elements elements of in present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..de336f68f08c856607d6ceaa98a3da5f92aaf690
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5d63caa644a8f8bcbc159304f865ef4c716b2588b6a360da81d41adac38715d
+size 190984
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..9186e3d4c4282cad041fb3a334dc9e6af15e7373
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is sitting on a rug in a living room, surrounded by two dogs. The man is wearing a blue shirt and has a surprised expression on his face. One of the dogs is wearing a hat and has a halo above its head, suggesting a humorous or playful tone. The living room is furnished with a couch and a coffee table, and there are pictures hanging on the wall. The overall style of the video is casual and light-hearted, with a focus on the interaction between the man and the dogs."
+    ],
+    "video_ids": [
+        "5vAL64qctfI_36_88to215"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, two dogs (one wearing a hat and having a halo).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting on the floor in a living room. There are two dogs visible: one is a German Shepherd with a digitally added hat and halo, and the other is a brown dog sniffing the floor. The man and the two dogs are clearly present, fulfilling the 'Object(s)' condition as described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting with the floor with a living room with He are two dogs with, one wearing wearing C Shepherd wearing a red added red and halo, and the other is a Golden dog withouting the German. The presence is the dogs dogs are the present, and the 'Object(s)' condition as described.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..549b88443eb7cc422c4ac07f1c7e74e8ae6d37eb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fdcf4028acb40d225df7c84bd92404b02426ae068c492591f4d945d04a9acb22
+size 190239
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..b3958b517a962454a384d2b354ce3d8798d71427
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a 3D animated orange tabby cat with large, expressive eyes and a furrowed brow, giving it a sad or concerned expression. The cat is standing on a green grassy surface with a few scattered leaves and flowers. In the final frame, the cat is holding a small black object, possibly a toy or a piece of food, in its paws. The style of the animation is realistic with attention to detail in the cat's fur and the environment. The lighting and shadows are well-rendered, creating a lifelike appearance. The cat's pose and expression suggest a sense of curiosity or anticipation."
+    ],
+    "video_ids": [
+        "JpVBU3sPAWs_14_0to119"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A 3D animated orange tabby cat, a small black object? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a 3D animated orange tabby cat with large, expressive eyes, which is the central focus. The cat is holding a small black object, likely a hat or similar item, in its paws. The description of the main objects is accurate and matches the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a 3D animated orange tabby cat standing large, expressive eyes, standing matches consistent primary object. The cat is holding a small black object in which a toy, a item, in its paws. The background matches the ' object in accurately, the the content content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7af7591665189029244acfe4297a9a79eb170cdb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:191aa7fdcc95649cbc9fb01b12d636432f659237bf361a3cbb1029ec6159d725
+size 195051
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..5f6090fd0333c68279a1bd2c47127dce16a21c1f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a vibrant and colorful scene of a traditional Chinese feast. The table, adorned with a red and gold tablecloth, is laden with a variety of dishes served in white and blue bowls. The dishes include dumplings, noodles, and a variety of vegetables and meats. The table is set with chopsticks and spoons, ready for the guests to enjoy the meal. The video captures the essence of a Chinese feast, with its emphasis on variety, color, and abundance. The style of the video is realistic, capturing the details of the food and the table setting with precision. The colors are vivid and the lighting is bright, highlighting the textures and colors of the food. The video is a feast for the eyes, capturing the essence of a traditional Chinese feast."
+    ],
+    "video_ids": [
+        "dzzHQIkl-FI_206_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dishes (dumplings, noodles, vegetables, meats) in white and blue bowls, chopsticks, spoons? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a table with various dishes including dumplings, noodles, vegetables, and meats, served in white and blue bowls. Chopsticks and spoons are also visible on the table. While there are additional elements like a roasted animal head and a tissue box, these do not contradict the core description of the dishes and utensils.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows dishes variety with various dishes, dumplings, noodles, and, and meats, all in white and blue bowls. Theresticks are aons are also visible, the table. The the are additional elements like a cup chicken and and a bowl box, these do not contradict the core description of the dishes and utensils mentioned}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..84b992ad36e1927b590119bf63e74a6505b1b645
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d1d644f578c06415680c6ed4fce020117919ca7df27734b1b18f77ccddd76ce7
+size 105893
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..ea2e6dc5ec39db4c254ac00115bfb0bd57f6a08b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a conversation between two men in a garage. The man on the left, wearing a gray jacket, is gesturing with his hands as he speaks, while the man on the right, dressed in a black jacket, listens attentively. They are standing in front of a blue motorcycle, which is parked on a red stand. The garage is filled with various items, including a shelf stocked with tools and a tire. The scene suggests a casual, friendly discussion, possibly about the motorcycle or other topics of interest. The overall atmosphere is relaxed and comfortable, indicative of a friendly gathering or a shared passion for motorcycles."
+    ],
+    "video_ids": [
+        "_pjbeC_YzUI_24_21to162"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, one in a gray jacket gesturing, the other in a black jacket listening. A blue motorcycle and a red stand.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men: one in a gray jacket who is gesturing while talking, and the other in a black jacket who is listening. A blue motorcycle is prominently visible in the foreground, and a red stand is also present under the motorcycle. All elements described in the condition are present and do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts two men in one in a gray jacket who appears gesturing, the, and the other in a black jacket who appears listening. The blue motorcycle with visible visible in the background, and a red stand is supporting present, the motorcycle. The elements in in the condition are present and match not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..62f5c2c738827d9b96d91582bf5f319442db7810
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8eb5fe5d50721c45b4d4c01a0fddee522d44555215fa3fc2b6f97f4433fc5ca2
+size 179481
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..e3dfbf0ca48b90fc6871c071bf829d414a890b6a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a city by the water at sunset. The sky is painted with hues of orange and pink, casting a warm glow over the cityscape. The city is nestled on the banks of a river, with boats gently bobbing on the water. The buildings, bathed in the soft light, reflect the colors of the sky.  Two people are seen standing on a cliff overlooking the city, their silhouettes adding a human element to the scene. They appear to be enjoying the view, perhaps taking a moment to appreciate the beauty of the city at sunset.  The video is shot from a high vantage point, providing a panoramic view of the city and the river. The camera seems to be moving along the cliff, giving the viewer a sense of movement and progression. The overall style of the video is calm and peaceful, capturing the tranquility of the city at sunset."
+    ],
+    "video_ids": [
+        "B5lfiDsPg3w_19_0to180"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two people standing on a cliff overlooking the city, with their silhouettes visible.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two people standing on a cliff overlooking a city, and their silhouettes are visible against the bright background of the water and sky. The presence of additional elements like the surrounding landscape and other people does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a sil standing on a cliff overlooking a city during with their silhouettes are clearly against the vibrant sunset of the sunset and sky. The scene of the elements like the city city and the buildings on not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..31ffb07e3a891c442829573e85dfc4fb011f63cf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1223-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:413aa35844f131fabd174400f3e0175663366ef3101411764130fec37b2164a4
+size 238231
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..e67be9e5104fae42aa664861dbbd734307368c73
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person's hand holding a blue and white figurine, which appears to be a dragon or a similar mythical creature. The figurine is intricately designed with detailed scales and a flowing mane. The background features a collection of similar figurines, each with unique designs and colors, displayed in a glass case. The style of the video is a close-up shot, focusing on the figurine in the hand, with the background slightly blurred to emphasize the figurine. The lighting in the video is bright, highlighting the colors and details of the figurine. The overall impression is that of a collector showcasing a prized piece from their collection."
+    ],
+    "video_ids": [
+        "TpAOvmaPMCc_5_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person's hand holding a blue and white dragon-like figurine, a collection of similar figurines in the glass case.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person's hand holding a blue and white dragon-like figurine, and in the background, there is a glass case filled with similar figurines, which matches the described condition. The presence of additional elements (like the person's wristwatch or the rotating base) does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person's hand holding a blue and white dragon-like figurine, which in the background, there is a glass case with with similar figurines. which matches the description '. The presence of the elements inthe the glass's hand and) the glass display of does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f62403dd4325e818ea949eb33afbf48a520ca02b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34964fe5db1ac3ef4c48d8881436b18dffeb60e6d69dd928223f74fceaa69b25
+size 119788
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..1e64ce0ea8981c768b46fae764f696f6e7892bcc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a baby girl in a stroller, dressed in a blue and pink floral dress, with a pink bow in her hair. The stroller is white with a blue and green patterned border. The baby is holding a cell phone in her hand. The stroller is placed on a concrete patio, with a grassy area and a rock garden in the background. The baby appears to be looking at the cell phone, possibly playing with it. The video is a simple, everyday scene, capturing a moment of a baby's life."
+    ],
+    "video_ids": [
+        "K1Y0rJWZxHU_11_0to126"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A baby girl in a stroller, a stroller, a cell phone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a baby girl in a stroller, which matches the core description. The stroller is clearly visible, and a cell phone is also present next to the baby. Additional elements like a person's legs in the background do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a baby girl sitting a stroller, which ful the ' description. The babyroller is visible visible, and the baby phone is being present in to the baby, The elements like the colorful's hand in the background do not contradict the core and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c0db9b6dc65b8d9d8a95e0d432ae3eb7844e66b9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6b0c476d0b2066a40b0521545e3350a19fa11bffc24ff61fdd2fa458d55248d
+size 305025
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..effba7135b66db478567507d11d2d9492f2e538c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a brown scooter parked on a gray floor in a showroom. The scooter has a sleek design with a white seat and a silver exhaust pipe. The handlebars are equipped with mirrors and a digital display. The scooter is positioned in front of a black wall with a curved archway. The lighting in the showroom is bright, highlighting the scooter's features. The overall style of the video is clean and modern, emphasizing the scooter's design and features."
+    ],
+    "video_ids": [
+        "Z5X7xBZpq-I_1_0to188"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A brown scooter with a white seat and a silver exhaust pipe, equipped with mirrors and a digital display.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a brown scooter with a white seat and a silver exhaust pipe, equipped with mirrors and a digital display. These features are clearly visible throughout the video, matching the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a scooter scooter with a white seat and a silver exhaust pipe. equipped with mirrors and a digital display. The elements match consistent visible in the video, matching the description provided.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..30d877b51dca41588fee2fd36ad090e105541336
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:741149a3b1d0708a4c9794e8ae8859719d600fdad4255c5d140d16b83a4f1fb8
+size 148700
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..0e5ef17a6ad592801ae24b97cddc38adba651ae1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a modern, open-concept living space that transitions seamlessly into a kitchen area. The living room features a white sofa adorned with blue and gray throw pillows, a wooden coffee table, and a rug with a floral pattern. A wooden bench is placed near the stairs, providing a cozy seating area. The kitchen area is equipped with stainless steel appliances, including a refrigerator and a microwave, and a sink. The space is illuminated by recessed lighting and pendant lights hanging from the ceiling. The overall style of the video is clean, contemporary, and inviting, highlighting the functionality and aesthetic appeal of the space."
+    ],
+    "video_ids": [
+        "A_07I8wj2qY_38_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: White sofa, blue and gray throw pillows, wooden coffee table, floral-patterned rug, wooden bench, stairs, stainless steel appliances (refrigerator, microwave), sink.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a white sofa with blue and gray throw pillows, a wooden coffee table, a floral-patterned rug, a wooden bench, stairs, and stainless steel appliances including a refrigerator and microwave, along with a sink. All specified elements are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as showcasing showing a white sofa with blue and gray throw pillows, a wooden coffee table, a floral-patterned rug, a wooden bench, stairs, and stainless steel appliances including a refrigerator and a. as with a sink. The the objects are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6c5b318e673026f9f35ad23d749d7c8f6b3af423
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb9d8023f18c297d8f7c7c7af0ecb623fa52dde7884e1d6f799b1f290bb47634
+size 84600
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..e3fa1112fb17723f57115e03a3c3a453e6008baa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the serene beauty of a tropical beach. The first frame shows a clear blue sky above the turquoise ocean, with a small island visible in the distance. The second frame shows the ocean gently lapping against the white sandy beach, with a few palm trees scattered along the shore. The third frame shows the ocean's surface shimmering under the sunlight, with a few small waves breaking near the shore. The overall style of the video is calm and peaceful, with a focus on the natural beauty of the beach and ocean."
+    ],
+    "video_ids": [
+        "-KufBNhw2oM_68_0to126"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Ocean, Sky, Island, Palm Trees? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing an ocean with turquoise waters, a blue sky, an island with lush greenery, and palm trees along the shoreline. All specified elements are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting an ocean with varying waters, a sky sky with an island, a greenery, and palm trees. the shoreline. The these elements are present and contribute depicted,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e6f7c9d370b150fa3b717de106905f503d4d7a28
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7926e6600e2677aa535828adb4a4742cab1a3a9819ab4ac0f8efb32d98877e2b
+size 166439
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..1468e14d5ab773b41a03ca533a9ea290b1fa828d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a culinary tutorial on how to make a pasta salad. The style of the video is a close-up, step-by-step demonstration, focusing on the ingredients and the process of preparing the dish. The main ingredients include pasta, cherry tomatoes, black olives, and a creamy dressing. The video captures the vibrant colors of the ingredients and the creamy texture of the dressing. The pasta is cooked and mixed with the other ingredients, and the dressing is added to create a delicious pasta salad. The video is likely to be informative and visually appealing, with a focus on the preparation and presentation of the dish."
+    ],
+    "video_ids": [
+        "AMaiv5ZMvbg_0_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Pasta, cherry tomatoes, black olives, creamy dressing? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a pasta salad with visible spiral pasta, chunks that resemble cherry tomatoes, black olives, and a creamy dressing coating the ingredients. All elements mentioned in the description are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close dish with cherry pasta pasta, cherry of appear cherry tomatoes, and olives, and a creamy dressing being the pasta. The the mentioned in the ' are present, clearly with the image content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5fa2432470b0b259593690eb3e797b115489f0d5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:045e15d481f2c4da5247310b101e1f7c13e4ce41ac9a1abc7f3757cc8341fb9a
+size 116362
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..ee045129a48d9c025b474d8724fb27a07a1845f5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a drone shot of a mountainous village with a castle on a cliff. The village is nestled in a valley with houses and buildings scattered across the landscape. The castle, perched on the cliff, overlooks the village and the surrounding mountains. The sky is clear with a few clouds, and the sun is shining brightly, casting shadows on the landscape. The drone flies over the village, providing a bird's eye view of the castle and the village below. The video captures the beauty of the landscape and the historical significance of the castle."
+    ],
+    "video_ids": [
+        "nVUdf718oLw_32_137to298"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Village, castle, houses/buildings, sky, clouds, sun? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a village nestled in a valley, a castle perched on a hilltop, numerous houses and buildings with red roofs, a bright blue sky with scattered clouds, and sunlight illuminating the landscape. All specified elements are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a village with in a valley, a castle perched on a hill,, numerous houses and buildings, red roofs, a vibrant blue sky, a clouds, and a illuminating the scene. The these elements are present and contribute depicted,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ab7fe0f3e223ab235cd8881831109816d59097af
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf3d9d00c6e77f9eb2012b9b4a4cb6b2b14dd0e5f80147457d0f59384d1063e2
+size 197816
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..0d938e67c7d84909f7e74410fb29e550da8861ec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a collection of freshly baked muffins cooling on a wire rack. The muffins are golden brown with a few dark spots, indicating they have been baked to a perfect texture. The wire rack is placed on a wooden table, which provides a warm and rustic backdrop to the scene. The muffins are arranged in a circular pattern, with each muffin slightly overlapping the next, creating an appealing visual effect. The video captures the essence of home baking, with the muffins looking delicious and inviting. The overall style of the video is simple and straightforward, focusing on the muffins and their presentation."
+    ],
+    "video_ids": [
+        "G8QA53P7MyU_17_0to172"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Freshly baked muffins? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows several golden-brown, rounded baked goods on a cooling rack, which are visually consistent with freshly baked muffins. The surface texture and appearance of the items match typical muffins, and there are no elements that contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a muff-brown muff glossy objects goods that a cooling rack, which are consistent consistent with freshly baked muffins. The texture appears and color suggest the baked suggest the muffins, and there are no elements in contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..33ae3f89a8c2b3a8744543ccbdf361d1e1df6f20
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1224-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:339ea4a17e9e9ca6e33498e9c486926fc60975d6b15ddcfa522cd42443aad1b9
+size 83813
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..2bacace516c232c2608229778273ec86b3c0bfb1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with long blonde hair, wearing a white baseball cap and a black jacket. She is seated in a room with a window covered by gray blinds. The woman appears to be looking off to the side, her expression neutral. The room has a simple, uncluttered appearance, with the blinds being the only notable feature in the background. The lighting in the room is soft and even, suggesting an indoor setting. The woman's attire and the casual setting suggest a relaxed, informal atmosphere."
+    ],
+    "video_ids": [
+        "CJ8MsaoY7Ic_0_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with long blonde hair, wearing a white baseball cap and a black jacket. She is seated and looking off to the side with a neutral expression.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with long blonde hair, wearing a white baseball cap and a black jacket, seated and looking off to the side with a neutral expression. These elements match the description provided in the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person with long blonde hair, wearing a white baseball cap and a black jacket. seated and looking off to the side with a neutral expression. The elements match the description provided, the question.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b6c2c8b3b8189fa80127853568c43ee5b2735126
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:56be397c608c43d055c9b82bfa3bc68d5f25eae872993cf23397dbd276654bb9
+size 164957
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..e7353ae8d6e5fb59d97dcf08302f3e0ae7cc2bf5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a silver Porsche Cayenne driving on a snowy road. The car is sleek and modern, with a shiny exterior that reflects the surrounding environment. The road is covered in a layer of snow, and the car's tires leave tracks as it moves. The car is driving towards the camera, and the angle of the shot suggests that the viewer is following the car as it moves down the road. The background is filled with trees and houses, indicating that the car is in a residential area. The overall style of the video is realistic and captures the beauty of the snowy landscape."
+    ],
+    "video_ids": [
+        "8mQDSR0m7-k_0_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A silver Porsche Cayenne? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a silver Porsche Cayenne, which matches the specified object. The car is prominently featured throughout the frames, and its distinctive design, including the Porsche emblem and model-specific features, is visible. The surrounding environment does not contradict the object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a silver Porsche Cayenne driving which matches the description object. The car is prominently displayed in the frames, and its design design, including the shape emblem and the-specific features, is visible. The snowy environment, not contradict the core description,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fe7a74b4154f100bb909ee4699a95422b91b6a6a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d245c5fc22f4bc58ca5d05633bed0f9cc1476b3524f540daf539994517a9e6da
+size 281546
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..1bd7a89c1b149a70d1a8c0257bb360c37f54ae75
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a purple shirt, who appears to be in the middle of a conversation or presentation. He is wearing glasses and has a serious expression on his face. Behind him, there is a bookshelf filled with various books, suggesting that he might be in a library or a study room. To the right of the bookshelf, there is a potted plant, adding a touch of greenery to the scene. The man's gestures and the context of the setting suggest that he is discussing a topic related to literature or academia. The overall style of the video is straightforward and professional, with a focus on the man and his surroundings."
+    ],
+    "video_ids": [
+        "1WS712DHfmg_2_0to122"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a purple shirt, wearing glasses, and a potted plant.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a purple shirt and glasses, and there is a potted plant visible in the background. These elements match the core description provided, and no contradictory elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing a purple shirt and glasses. and there is a potted plant visible in the background. The elements match the description description provided, and the additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6748c180157ed41a3e815f51b0d37673a8a0340f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13d20eaed93ce21fd44b7b1369e6f419628df0f62c78f4ea8e230b832907a27f
+size 109357
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..703a8c6be4c7def1ec260699a79ea16fd395edac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a green grasshopper in its natural habitat. The grasshopper, with its vibrant green body and brown legs, is seen perched on a blade of green grass. The grasshopper's antennae are extended, and its eyes are focused, suggesting it is alert and aware of its surroundings. The background is a blur of green, indicating a dense grassy area. The video is shot in a realistic style, with a focus on the grasshopper and its immediate environment. The camera angle is slightly above the grasshopper, providing a clear view of the insect and its surroundings. The video does not contain any text or additional objects. The overall impression is of a peaceful moment in nature, with the grasshopper as the main subject."
+    ],
+    "video_ids": [
+        "BvVChvKd46M_17_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A green grasshopper? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a green grasshopper clinging to a blade of grass, which matches the description. The grasshopper is clearly visible and is the central focus of the video, with no elements contradicting the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a close grasshopper with to a blade of grass. which align the description of The vibranthopper is the the with the the main focus of the video. fulfilling its conflicting thating the core.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a2a20ee92cacba5b88633b5fd58eec3cc26a813e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a07ca8007a65c37a8e9cb8007690b0c43e3de515322b4b11a86cff5de52c3707
+size 130495
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..d8364e5dae4573a01e7edaee1f219a5c6adc93ca
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a sandy environment, likely a desert or a similar arid area. The sand is a mix of light beige and darker brown hues, indicating varying particle sizes and possibly different mineral compositions. A small ant is seen moving across the surface of the sand, its reddish-brown body contrasting with the lighter sand. The ant's legs are clearly visible as it scurries along, suggesting a sense of urgency or purpose in its movement. As the video progresses, the ant continues to traverse the sandy terrain, leaving behind a trail of disturbed sand particles. The camera remains stationary throughout the sequence, focusing on the ant's journey across the sand, providing a detailed and intimate look at the insect's behavior in its natural habitat. The BBC logo in the top left corner suggests that this footage is part of a documentary or educational program."
+    ],
+    "video_ids": [
+        "02280eea91cae67ddf9f213328c13f8415c6e37bb1bd475780e3cd0ffa1c29f9"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small ant with a reddish-brown body.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small ant with a reddish-brown body, which matches the description. The ant is clearly visible against the sandy background, and its coloration and size are consistent with the specified condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a close ant with a reddish-brown body, which is the description. The ant is visible visible in the background background, and its size is is size are consistent with the description '.\"\n}<|im_end|>\n addCriterion",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..30c83658a3b3a441767574c8a3f5e9726b1ac8bd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:997705882fe7990722ee49774df564302126808c74b935ac777f0674f1522d5d
+size 176945
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..d085ff954a9090d4da0c520342c3d808d93a9776
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a group of six friends are enjoying a sunny day outdoors. They are seated on a bench, with the camera capturing their cheerful expressions and playful poses. The friends are dressed in casual, colorful clothing, and they are all smiling and looking directly at the camera. One of the friends is holding a smartphone, presumably taking a selfie of the group. The background features a lush green park with trees and grass, providing a serene and natural setting for their gathering. The overall style of the video is light-hearted and fun, capturing a moment of joy and camaraderie among the friends."
+    ],
+    "video_ids": [
+        "aB7s0-unP_Y_19_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Six friends seated on a bench, one friend holding a smartphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows six friends seated on a bench, and one of them is holding a smartphone to take a selfie. This matches the 'Object(s)' condition described. The background and additional elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a individuals seated on a bench, and one of them appears holding a smartphone, take a selfie. The matches the descriptionObject(s)' condition as in The friends and the elements, not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fc2d4ff606d28920f5f3eea46d0fe57cf80ec395
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0db97644eee8798f33a93d3756316afd14347ae24bf14e2220853c4decf2e40f
+size 225427
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..2356166860c54e94a0719f3febc7267e8a22f266
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a thrilling moment at a snowboarding event. The crowd of spectators, bundled up in winter gear, watches in anticipation as the snowboarder takes off from the ramp. The snowboarder, dressed in a vibrant red jacket, is captured mid-air, performing an impressive trick. The snowboarder's body is angled towards the ground, suggesting a high level of skill and control. The snow-covered slope and the clear blue sky in the background add to the excitement of the scene. The video is shot from a high angle, providing a bird's eye view of the action, and the camera follows the snowboarder's trajectory, creating a sense of motion and excitement. The overall style of the video is dynamic and action-packed, capturing the adrenaline-fueled spirit of the sport."
+    ],
+    "video_ids": [
+        "AaZ5grifocc_0_0to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Snowboarder, crowd of spectators? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a snowboarder descending a snow-covered ramp and a large crowd of spectators watching the event. These elements align with the 'Object(s)' condition described, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a snowboarder performing a snowy-covered slope, a crowd crowd of spectators watching the event. The elements directly with the 'Object(s)' condition,, making there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0addfc27e826d9c292443e71f72e9873d7e8505c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a2663faed4732ddd9c79df8a5128ffa030cdb6ae0605bec7dce0492abd897bb
+size 250516
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..62ac1853cf6c682d9e874fb973aa7c172e357299
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two women are standing in front of a red pickup truck on a dirt road. The woman on the left is wearing a green shirt and blue jeans, while the woman on the right is wearing a red shirt and blue jeans. They are both smiling and appear to be in a good mood. The truck is parked on the side of the road, and there is a fence in the background. The sky is overcast, and the ground is covered in grass. The overall style of the video is casual and friendly, with a focus on the interaction between the two women and their surroundings."
+    ],
+    "video_ids": [
+        "EO273WqQ2Qo_11_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two women, one in a green shirt and blue jeans, the other in a red shirt and blue jeans.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two women standing in front of a red truck on a dirt road. The woman on the left is wearing a green shirt and blue jeans, and the woman on the right is wearing a red patterned shirt and blue jeans. This matches the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two women standing in front of a red truck. a dirt path. One woman on the left is wearing a green shirt and blue jeans, while the woman on the right is wearing a red shirted shirt and blue jeans. The matches the description provided, the questionObject(s)' condition.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3ee574f92954924db84f8e149b484009b22c3bce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1225-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b17b756651e892618cd7696dfc46b56e14d1bf130d87c4c17bb680c3162a4bd0
+size 233615
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..b5966eac0b734ff7d2d1568aacab18eb1d77f856
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man with white hair and a blue shirt is engaged in a conversation with a bald man in a police uniform. The setting appears to be an indoor space, possibly a garage or a storage area, as indicated by the presence of a large truck in the background. The man in the blue shirt is seated, while the police officer stands, suggesting a formal or official interaction. The lighting in the scene is artificial, casting shadows and highlighting the contours of the men's faces. The overall style of the video is realistic, capturing a moment of dialogue between two individuals in a non-fiction setting."
+    ],
+    "video_ids": [
+        "bBaWlF0ME-w_6_68to214"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with white hair and a blue shirt, a bald man in a police uniform.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men: one with white hair wearing a blue denim shirt, and another who is bald and dressed in a police uniform with \\\"Los Angeles Police\\\" visible on his sleeve. These elements match the description exactly.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two men: one with white hair wearing a blue shirt shirt and and the who appears bald and dressed in a police uniform. aPOL Angeles\\\" Department written on the shirt. The elements match the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9958b01b3f41b7784a180ab0c23dfbd2fcc203b3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:679e29112e5ab44c6245407e3dd37b29184441f0de082a23442e6db37ca87751
+size 124800
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..78472cea80ea90f9b8cf40010461a35a01ff80e6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated scene featuring two characters, a green alien-like creature and an orange monster, sitting at a desk in a classroom setting. The green creature is holding a pencil and appears to be writing or drawing on a piece of paper. The orange monster is smiling and has its arms outstretched, as if it is excited or enthusiastic. The classroom has wooden desks and chairs, and there are bookshelves in the background filled with books. The overall style of the video is colorful and cartoonish, with a focus on the characters' exaggerated facial expressions and body language."
+    ],
+    "video_ids": [
+        "xHHniI1WmZs_53_0to120"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two characters: a green alien-like creature and an orange monster.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly features two main characters: a green alien-like creature with large eyes and brown hair tied with orange buns, and an orange monster with a single large eye, wearing a leopard-print outfit. Both characters are prominently displayed in the classroom setting, matching the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts two distinct characters: a green alien-like creature and a eyes and a hands, in a ribuns, and an orange monster with a large horn eye and a a red print shirt. Both characters are positioned displayed in the scene setting, fulfilling the description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7f6ad9bb5c9d6b1c5db9b5f609aea9d6269e56e7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4aaa93622a2a4f9e593d8e59afed6ac4a45cb97621ef063cec7058dfb3ee5dbc
+size 130564
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..0ebb57432bd64fb89194fec2381908e1a71cf85f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment on a football field where a coach is giving instructions to his players. The coach, wearing a headset, is in the center of the frame, surrounded by his players. He is gesturing with his hands, emphasizing his point. The players are dressed in their team uniforms, indicating that they are part of the same team. The background is filled with the blurred figures of other players and coaches, suggesting that this is a busy and active scene. The focus is clearly on the coach and his interaction with his players, indicating that this is a crucial moment in the game. The style of the video is realistic, capturing the intensity and focus of the coach and his team."
+    ],
+    "video_ids": [
+        "FlLVDlZbbEE_5_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Coach, players? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a coach wearing a headset and a team jacket with NFL and Raiders logos, actively communicating with an open mouth, which aligns with the 'Coach' object condition. Additionally, players in team uniforms are visible in the background, fulfilling the 'players' condition. The scene is consistent with an NFL football game setting.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a coach wearing a headset and a player of, a logos team logos, indicating engaging with players NFL mouth, which suggestss with the 'Coach' condition.. Additionally, there in light uniforms are visible, the foreground, fulfilling the 'Players' condition. The presence is set with a NFL game game setting,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ac9201452cdd35ccf502fc625ed09b6c0193153a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03e480811c614dc26f0000bd192cc41e20782cc9880cf0388581f6dca89de403
+size 144167
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..f78ac0aba8b4a1b523e446fa20e483dcb82d9a0b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene mountain landscape, with a wooden pole standing prominently in the foreground. The pole, which appears to be a marker or post, is surrounded by a small pile of rocks, adding a rustic charm to the scene. The background is dominated by majestic mountains, their peaks reaching towards the clear blue sky. A winding dirt path cuts through the grassy field, inviting viewers to imagine a journey through this picturesque setting. The overall style of the video is naturalistic, with a focus on the beauty of the landscape and the simplicity of the wooden pole and rock pile. The colors are vibrant, with the green of the grass and the blue of the sky providing a striking contrast to the brown of the mountains and the wooden pole. The video is likely to evoke feelings of tranquility and awe in the viewer, as they take in the grandeur of the mountains and the peacefulness of the field."
+    ],
+    "video_ids": [
+        "zpHtMxqoCWk_74_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A wooden pole in the foreground, surrounded by a small pile of rocks; a winding dirt path through the grassy field.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a wooden pole in the foreground, embedded in a small pile of rocks, and a winding dirt path through the grassy field. These elements are consistent with the description, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a wooden pole in the foreground, surrounded in a small pile of rocks. and a winding dirt path through a grassy field. The elements match consistent with the description provided and there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..86cfc9571e11538cae2aeaaeee047aadeeb77e98
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a8cfd6003ca37a73ace64cd889edba56f8296fb6f4b85aa94ac2c8f52a22ac1b
+size 174747
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..8c935f9c4f7970389d203d6a055e3c97702bbb6a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of two pieces of chicken being cooked in a pot of boiling water. The chicken pieces are partially submerged in the bubbling liquid, which is filled with chunks of white onion floating around them. The water is vigorously boiling, creating a dynamic scene with bubbles rising to the surface. The chicken appears to be seasoned, with visible specks of black pepper on its surface. The pot itself has a metallic appearance, with a handle visible on the left side, suggesting it's placed on a stovetop. The focus remains on the cooking process, highlighting the transformation of the ingredients as they cook. There are no significant changes or movements throughout the frames, maintaining a steady and continuous depiction of the boiling chicken and onions."
+    ],
+    "video_ids": [
+        "a40b109b420d2213e9f67a50f6b191ffd5ae39e17a25cf87bc99a594ad93990b"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two pieces of chicken, chunks of white onion, boiling water, a metallic pot.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two pieces of chicken, chunks of white onion, boiling water, and a metallic pot. All elements described in the condition are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two pieces of chicken, chunks of white onion, and water, and a metallic pot. The the described in the condition are present in match depicted in the video.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ad41eb9a5bb7a327c99adc2e82003cadfa981c31
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23d10aa4cbe8c2388c53c8bf702942f790df75cbb407b21daa15fc7430fb3ad3
+size 173939
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..e530ed15eaabab12d24e0240235fc513d0cf8427
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated sequence featuring a doll with blonde hair and blue eyes. She is wearing a gray sweater with a pink heart on it. The doll is standing in a kitchen setting, with a sink to her left and a counter with various items, including a bowl and a vase, to her right. The kitchen has a vintage aesthetic, with white cabinets and a marble countertop. The doll is looking directly at the camera, and her expression is neutral. The lighting in the scene is soft and warm, creating a cozy atmosphere. The style of the video is realistic with a touch of whimsy, and it appears to be aimed at a younger audience."
+    ],
+    "video_ids": [
+        "ve9jK-FXDS8_15_68to204"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A doll with blonde hair and blue eyes, wearing a gray sweater with a pink heart.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a doll with blonde hair and blue eyes, wearing a gray sweater with a pink heart design, which matches the description. The background elements, such as the kitchen shelves and plants, do not contradict the core description of the doll.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a doll with blonde hair and blue eyes, wearing a gray sweater with a pink heart.. which matches the description provided The setting appears, such as the kitchen setting and items, do not contradict the core description of the doll.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..113d01170b43732f27402f19987fa8c9a58e5143
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf9595f2161dfe8e24753323d688c79aa4e3a5dde5e461e41cdfccd3af94a6f6
+size 276963
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..2ed56ed24ec48930932de6f323638727234c6098
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a yellow sports car with red rims, showcasing its sleek design and vibrant colors. The car is parked in a lot, with other cars visible in the background. The camera angle is low, emphasizing the car's aerodynamic shape and the contrast between the yellow body and red rims. The car's design suggests speed and luxury, with a focus on performance and style. The video is likely intended for promotional purposes, highlighting the car's features and appeal to potential buyers or enthusiasts."
+    ],
+    "video_ids": [
+        "IoTVHwF2feM_9_0to104"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A yellow sports car with red rims.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a yellow sports car with red rims, which matches the core description. Additional cars in the background do not contradict this primary focus.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a yellow sports car with red rims, which matches the description description. The elements and the background do not contradict the main focus.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e900ca77e10ac678aa48ff6282e2597e0cffdbc0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b923211a3e652f57266c88f42852698050063366d8df033da4d027341e9d9e14
+size 156964
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..0648d6d25eadd7ac874895a372b20c2554017759
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a car in a garage, being worked on. The car is a silver sports car with a black interior. The car is missing its hood and front bumper, revealing the engine and other components. The car is parked on a red floor, and there are other cars and motorcycles in the background. The garage is well-lit, and there are tools and equipment scattered around. The style of the video is a time-lapse, showing the car being worked on over time. The video captures the process of the car being taken apart and put back together."
+    ],
+    "video_ids": [
+        "n4ja2HSn6JQ_77_99to225"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A silver sports car with a black interior, other cars and motorcycles? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a silver sports car with a black interior, as described. Other cars are visible in the background, and while motorcycles are not explicitly visible, the presence of other vehicles and garage equipment does not contradict the description. The core elements of the description are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a silver sports car with a black interior, which described. The cars and visible in the background, and there there are not explicitly mentioned, the presence of other vehicles in the-like suggests not contradict the description of The video elements of the video are met represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dddd890977f7283e7ed2d0aa0cb71ad562c52b89
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1226-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d5dde2b366e75817ba3e28adea69ec1c7463b038172aed1eb68827f9c8c956a
+size 181954
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..c098bcee29a96cdfdc0e5c5ca69bdff0089796ca
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a blue SUV parked in a parking lot. The car is positioned at an angle to the camera, allowing a clear view of its side profile. The vehicle is equipped with a set of silver rims and has a distinctive rear tail light design. The parking lot appears to be empty, with the SUV being the only vehicle visible. In the background, there are trees and a clear blue sky, suggesting a sunny day. The style of the video is a straightforward, real-life depiction with no additional effects or filters applied. The focus is on the SUV, and the background is kept simple to avoid distractions."
+    ],
+    "video_ids": [
+        "8BjoGG6aCiQ_10_78to241"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue SUV with silver rims and a distinctive rear tail light design.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a blue SUV with silver rims and a distinctive rear tail light design, matching the description. The vehicle's color, wheel design, and tail light styling are clearly visible and consistent with the specified criteria.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a blue SUV with silver rims, a distinctive rear tail light design, which the description provided The vehicle is color, the design, and tail light shape are clearly visible and consistent with the given object.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..47cbce350fb8267931556a99bebc12a4bf823f18
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76a1611ba6571b2d26e64e372ef30a642ab9ccd91bfc01fdd69fbe1ccec7bf08
+size 272264
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..505d43f78664b5eb3fab406e0f2fb31ef5d3fe7f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a montage of a dirt bike being ridden on a street. The first frame shows the dirt bike parked in front of a garage door. The second frame shows the dirt bike being ridden down the street. The third frame shows the dirt bike parked in front of a building. The style of the video is a time-lapse montage, capturing the progression of the dirt bike from being parked to being ridden and then parked again. The video is set in an urban environment, with buildings and a street visible in the background. The dirt bike is the main focus of the video, with its orange and white color standing out against the urban backdrop."
+    ],
+    "video_ids": [
+        "NVSH9Z7vPrs_55_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dirt bike? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a dirt bike, specifically a KTM model, which is clearly visible in the central image. The bike's design, including knobby tires, high suspension, and off-road styling, matches the description of a dirt bike. The surrounding elements (e.g., street view, mirrors) do not contradict this core object.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a dirt bike, which a KTM model, which is clearly visible in both frames part. The bike is design, including itsobby tires and a ground, and off-road styling, is the typical of a dirt bike. The video environment,a.g., the,, buildings) do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fa361df2606215015f8abd9f8fd13cfac539c248
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b8be8d2f9f058f0097e1c4babf2667f0369d286fd779488ea6f7a309e99592d
+size 300020
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..a159a0e1a86437f926cbca69ae07bc2a30710d82
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the beauty of a pink flower in bloom, set against a backdrop of green foliage. The flower, with its vibrant pink petals, is the main focus of the video, drawing the viewer's attention with its delicate beauty. The green foliage provides a natural and contrasting backdrop, allowing the pink flower to stand out. The video is shot in a close-up style, allowing the viewer to appreciate the intricate details of the flower. The pink flower is in full bloom, indicating that the video was likely shot during the spring or summer months. The overall style of the video is simple yet elegant, focusing on the natural beauty of the flower and its surroundings."
+    ],
+    "video_ids": [
+        "no8Y4AjMhps_3_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A pink flower in full bloom with vibrant pink petals.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows pink flowers with vibrant pink petals, appearing in full bloom. Although the focus is shallow and some parts are blurred, the visible portions clearly match the description of pink flowers in bloom. Additional elements like green stems and background foliage do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a flowers with vibrant pink petals, which to full bloom. The there video is on, the petals of blurred, the core petals of match the description of a flowers in full. The green like the leaves and leaves foliage are not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d8d08d1d266a779c1e577826cb8aa1201d6f17c6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1f10b01a387e07d86a478b2ab1e92b7756c9a391d07c26831354596c876bccf4
+size 119910
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..f6ba04d8dacc691f2f56154a3cbc1e29c7394be6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit and tie, sitting in a chair and speaking. The man appears to be middle-aged with short, light-colored hair. He is wearing a dark suit with a white shirt and a patterned tie. The background is plain and light-colored, providing a neutral backdrop that focuses attention on the man. The style of the video is straightforward and professional, likely intended for a business or informational context. The man's expression is serious, suggesting that he is discussing a topic of importance or providing information. The lighting in the room is even, with no harsh shadows, which contributes to the professional appearance of the video."
+    ],
+    "video_ids": [
+        "Sdl-rppq-fY_13_0to188"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, sitting in a chair and speaking.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed in a suit and tie, seated in a chair, and actively speaking. The background and logo do not contradict this core description, and the subject's actions align with the requirement.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in in a suit and tie, seated in a chair, and appears speaking. The attire is attire suggest not contradict the description description, and the presence's attire align with the ' of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..938e760ac76674af6c28b5886ad75fb09b12cfa6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:295a1cd3e8acef1013928c028e8040de4840a2669cf39a69bc813fdb3e1abd64
+size 216263
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..82744d7df002d8e9d2b101f33692f6b5674581f3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two women are engaged in a conversation in a kitchen. The woman on the left, wearing a green shirt, is speaking with a concerned expression. She is gesturing with her hands, indicating an active discussion. The woman on the right, dressed in a black and white plaid shirt, is listening attentively. The kitchen setting is warm and inviting, with wooden cabinets and a window that lets in natural light. A vase with flowers adds a touch of color to the scene. The overall style of the video suggests a casual, intimate conversation between friends or family members."
+    ],
+    "video_ids": [
+        "WkrfJhVYuvE_3_0to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two women, one in a green shirt and one in a black and white plaid shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two women standing in a kitchen. One woman is wearing a green shirt, and the other is wearing a black and white plaid shirt, matching the description. Additional elements like the kitchen background and decorations do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two women, in a room-like One woman is wearing a green shirt, and the other is wearing a black and white plaid shirt. which the description provided The elements like the kitchen setting and the do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a762fe43c07cefe68b0e0211d1986a439329c425
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d1c586f6e496b5eac986c9b2721c62cdb7324c07dedf24673322f0b1a9e063e6
+size 167526
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1d359ea011578f35b7e53044861fdb586500fd1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a garage, sitting on a yellow box. He is wearing a black t-shirt and a black baseball cap. The garage is filled with various items, including a motorcycle, a trash can, and a suitcase. The man appears to be in the middle of a project, as he is surrounded by tools and equipment. The garage has a rustic feel to it, with its concrete floor and metal walls. The man seems to be focused on his work, as he is looking directly at the camera. The overall style of the video is casual and candid, capturing a moment in the man's day-to-day life."
+    ],
+    "video_ids": [
+        "K9ukHtJH_8I_25_0to133"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a yellow box, a motorcycle, a trash can, a suitcase, and tools/equipment.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting on the floor in a cluttered garage or workshop. A yellow box (tool kit) is visible on the floor near him. A motorcycle is partially visible in the background, along with a trash can and what appears to be a suitcase or storage container. Various tools and equipment are scattered around, including a tool box and other mechanical parts. All specified objects are present and do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting on a floor in a roomed room or workshop. He yellow box islikely box) is visible in the floor next him. A motorcycle is parked visible in the background. and with various trash can and various appears to be a suitcase. storage container. Various tools and equipment are scattered around, fulfilling what hammer box and what items items. The the objects are present, the not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..500b37cbf1bf823e69f6f84d6c9318b592f5e79b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:14f65371572995a0539dc145c2b1ac45185cdf3dada402feb53d67f1d6c53933
+size 224821
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..13f7d919b4eb93c6d76409a072dcb576d9f9ebf4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a cartoon character with a blue and white face, wearing a black and orange outfit, is sitting at a wooden table. The character is holding a plate of green food, possibly salad, and there is a bottle of ketchup next to the plate. The character appears to be enjoying the meal. The background features a green cabinet with a red and white object inside, and a wooden door. The overall style of the video is cartoonish and colorful, with a focus on the character and the meal."
+    ],
+    "video_ids": [
+        "tNJvUmCtLWk_244_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A cartoon character, a plate of green food, and a bottle of ketchup.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a cartoon character (a vampire-like figure) sitting at a table with a plate of green food (appearing to be spinach or similar greens) and a bottle of ketchup (red bottle with tomato imagery). All core objects mentioned in the condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a cartoon character witha blue-like figure with sitting at a table with a plate of green food andwhicharing to be vegetables or a)) and a bottle of ketchup.red in with a design). The the elements are in the condition are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b616d816b7c390f1216fb89bf6fdd2b94ea3a467
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ff60841fdda418253666d044da88da388ec53f0e73a784829ff7dfcd9f25ffd
+size 142999
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..adbb931d0482538f3e2e05c2a712fdab66e88512
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young girl dressed in a white lab coat, standing in front of a chalkboard filled with mathematical equations. She has a bow in her hair and is looking off to the side with a thoughtful expression. The style of the video is educational and informative, likely aimed at teaching viewers about mathematics or science. The girl appears to be the main subject of the video, with the chalkboard serving as a backdrop to emphasize the learning environment. The video may include voiceover narration or on-screen text to explain the concepts being displayed on the chalkboard."
+    ],
+    "video_ids": [
+        "goIvD4H8r2I_11_89to238"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young girl, a chalkboard filled with mathematical equations? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a young girl wearing a lab coat and a bow in her hair, standing in front of a background that resembles a chalkboard filled with mathematical equations and diagrams. The core elements described are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a young girl wearing a lab coat, standing bow in her hair, standing in front of a chalk that includes a chalkboard filled with mathematical equations. symbols. The presence elements of in present, match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7bada19b2958559fbbd5cf52ee6f78a35618a027
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1227-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:292644ed35a3b4d721beb79af338f999a3b902634d847ea580a22e1dcefb2774
+size 87370
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..55b40f30d8fb4dbe80ec44bf03ef93c7a51a3932
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a blue bowl filled with a colorful and appetizing meal. The meal consists of chunks of meat, possibly chicken, and a variety of vegetables including potatoes, carrots, and red peppers. The vegetables are cooked to perfection, with the potatoes appearing soft and the carrots and peppers maintaining their vibrant colors. The meat is tender and well-cooked, suggesting it has been slow-cooked or braised. The bowl is placed on a granite countertop, which adds a touch of elegance to the scene. The overall style of the video is simple yet appealing, focusing on the food and its presentation rather than any additional elements or distractions. The video likely aims to showcase the meal's deliciousness and the care that went into its preparation."
+    ],
+    "video_ids": [
+        "6bOZdNsGVHw_10_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Blue bowl, chunks of meat (possibly chicken), potatoes, carrots, red peppers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a blue bowl containing chunks of meat (likely chicken), potatoes, and red peppers. While carrots are not clearly visible, the core elements described \u2014 blue bowl, meat, potatoes, and red peppers \u2014 are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a blue bowl containing chunks of what,possibly chicken), potatoes, and red peppers. The the are not visible visible, the presence elements of in a bowl, meat, potatoes, and red peppers \u2014 are present. match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f6cc920da74e6d048bb00d70bfc1f3a921ab4a73
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ccf21c42ba6a5b3b516deb64ba3788112548b812539f466b5807ecbf2999971b
+size 115208
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..7ea1eb2f352795f1af7e7d131357adbb0d43f820
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a red car with black rims on a metal platform. The car is stationary, and the camera angle is from the side, focusing on the front wheel. The car's body is shiny, reflecting the light, and the rims have a glossy finish. The metal platform has a grid-like pattern, and the car is positioned on it, suggesting that it might be on display or being showcased. The style of the video is straightforward and clear, with no additional elements or distractions. The focus is solely on the car and its details, providing a clear view of the vehicle's design and features."
+    ],
+    "video_ids": [
+        "YZ-gmw913b8_10_22to197"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red car with black rims? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red car with black rims in the second frame, which matches the description. The car is positioned on a metal platform, and its black rims are visible. Although other cars and elements are present, they do not contradict the core description of the red car with black rims.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a red car with black rims. a foreground frame. which matches the description provided The first is positioned on a platform platform, and the vibrant rims are prominently, The the elements are a are present in they do not contradict the main description of the red car with black rims.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bc3fdc0a8a111a5a7295f11874a28341d284c37c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:84bb3ca75a3ad964a194db535ba8f6306817ae71865152da1596f9a4e74f69ed
+size 165352
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..f3fc6184eaeaf4c2fb7140d5ef82e1bab56584e5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two men are sitting at a wooden table in an outdoor caf\u00e9 setting. They are engaged in a conversation, with one man gesturing as he speaks. Both men are wearing black jackets with white logos on them. On the table, there are two cups of coffee, suggesting that they are enjoying a casual meeting over drinks. The background features a city street with buildings and potted plants, indicating an urban environment. The overall style of the video is casual and relaxed, capturing a moment of conversation between two individuals in a public space."
+    ],
+    "video_ids": [
+        "Z9JgpTV_bPc_10_16to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, two cups of coffee, city street, buildings, potted plants? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men sitting at a table with two cups of coffee, set against a backdrop of a city street with buildings and potted plants. All elements mentioned in the condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two men sitting at a wooden with two cups of coffee in which against a backdrop of a city street with buildings and potted plants. The the of in the condition are present and accurately depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..463687a830aa64af1cea0107fa6c74ebe06fe7c4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f24055926b7d76f04b9952594dbfa95387e6acfbcbeba1cd3b8aacb073fddeb
+size 174676
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..d64a8da78845e6e7356d67849603394f44e4f619
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a weather report. The weatherman, dressed in a suit and tie, is standing in front of a blue background with a weather map. The map displays the current temperature and dew point for various locations, including Litchfield, Milford, Storrs, and Hartford. The weatherman is smiling and appears to be in a good mood. The overall style of the video is professional and informative, typical of a news broadcast."
+    ],
+    "video_ids": [
+        "kg5FBBjHxOE_5_21to178"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A weatherman dressed in a suit and tie.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed in a suit and tie, standing in front of a weather graphic. He appears to be a weatherman presenting neighborhood weather data, which matches the core description. Additional elements like the weather graphic and station names do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person dressed in a suit and tie, which in front of a blue map background The appears to be presenting weatherman, a weather updates, which align the description description of The elements like the weather graphic and text information do not contradict the main but}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0d07682dabb78656ddd73f40c4842fe871be9639
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52d2169a38e5518dd7c0570f80ccf7771fab995997653cb35641811b1bb5e6ad
+size 179835
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..02e01f4683d1f2e0564eb981bc02bf373133ebf9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a promotional advertisement for Lamborghini, featuring a close-up of the brand's logo. The logo is a shield-shaped emblem with a gold bull in the center, set against a black background. The word \"Lamborghini\" is prominently displayed in large, bold, white letters. The style of the video is sleek and luxurious, with a focus on the brand's iconic logo. The background is a dark, textured surface that contrasts with the bright, metallic colors of the logo. The overall impression is one of high-end luxury and exclusivity."
+    ],
+    "video_ids": [
+        "hH6y48u2Afo_8_0to148"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Lamborghini logo (shield-shaped with a gold bull), 'Lamborghini' text in white letters.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully displays the Lamborghini logo, which is shield-shaped with a gold bull, and the text 'LAMBORGHINI' in white letters. These elements are clearly visible against a textured black background, fulfilling the specified conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a Lamborghini logo, which is shield-shaped with a gold bull in and the text 'LambORGHINI' in white letters. The elements are clearly visible and a dark background background, which the ' conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..12a16ec889c40efdadde7ce4a376f6b3022d94fa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c5ed6e56d861ddaf832f3d87c7401e9a19328f046ff869cab7564518b942dbf4
+size 211044
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..4003fc4811ecc105855a19023224b04cc9b7a253
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young man with glasses, wearing a blue denim shirt and a white baseball cap with a red and black design. He is seated in a room with a brick wall and a bicycle hanging on it. The man is gesturing with his right hand, which is blurred due to motion. His facial expression suggests he is in the middle of speaking or explaining something. The style of the video is casual and informal, with a focus on the man's actions and expressions. The room's interior suggests a personal or home setting."
+    ],
+    "video_ids": [
+        "7FfBzAMMZ8o_42_1018to1167"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man with glasses, wearing a blue denim shirt and a white baseball cap with a red and black design.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man wearing glasses, a blue denim shirt, and a white baseball cap with a red and black design, which matches the description. Additional elements like the brick wall and bicycle in the background do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man wearing a, a blue denim shirt, and a white baseball cap with a red and black design. which matches the description provided The elements like the background wall and the in the background do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..27bea320eb2e66401e6d8c5b68874f6f2c44df59
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:abe23c1a056748abbfe90a10c049d12e02fd00374f4925f9e747507bda1f4c88
+size 279876
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..fcaf64ce92d717260d65f23deed8deb97c2c3857
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a stop-motion animation set in a Halloween-themed dollhouse. The dollhouse is purple and white with a purple door and black windows. The scene is decorated with Halloween decorations such as spider webs, pumpkins, and skeletons. The dolls are dressed in costumes and are interacting with each other and the environment. The lighting is dim with string lights adding a festive atmosphere. The style of the video is playful and whimsical, capturing the fun and creativity of Halloween celebrations."
+    ],
+    "video_ids": [
+        "T2oEHGOpKZQ_82_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dolls dressed in costumes? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple dolls dressed in various Halloween costumes, such as witches, ghosts, and other spooky outfits, which matches the 'Object(s)' condition. The costumes are clearly visible and are the central focus of the doll arrangement in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three dolls dressed in costumes costumes costumes, which as a and a, and a spooky attire. which align the descriptionObject(s)' condition of The setting are clearly visible and the the main focus of the scene characters.\"\n the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..88fda38daee0bdf6152a4825ec9be70dbb27e966
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d76ae252d854e756b286aa59cff65e105c4c899cddfd2d438e50cd4c8782ceed
+size 155787
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..e9a9409566082761b58b98ef3852f537f9e4f07c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a spacious and well-lit nutrition room with a warm and inviting atmosphere. The room features a large dining table surrounded by blue chairs, providing ample seating for group meals or discussions. The kitchen area is equipped with modern appliances, including a refrigerator, oven, and sink, and is complemented by wooden cabinets and countertops. The room also has a window that allows natural light to flood in, creating a bright and welcoming environment. The overall style of the video is clean and professional, with a focus on the room's layout and design."
+    ],
+    "video_ids": [
+        "czs_Lmewp4I_8_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Large dining table, blue chairs, modern appliances (refrigerator, oven, sink), wooden cabinets, countertops, window? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a large dining table with blue chairs, modern appliances including a refrigerator, oven, and sink, wooden cabinets, countertops, and a window (visible as a glass partition with blinds). All elements described in the condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a large dining table with blue chairs, modern appliances including a refrigerator, oven, and sink, wooden cabinets, countertops, and a window.though in a sk panel). a). The the match in the condition are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5a6718afc5f861fecd1f9b8198b47bfb3b8910bd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1228-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7086791fdde09dd9a114ded580a6eb5c3da0c2f67bb9e4937ad756297aa1ac90
+size 126529
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..62d2f59a8f0ca930a3c63189dafcbe0c01940d9a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man giving a speech or presentation. He is wearing a suit and tie, and has a microphone attached to his ear, suggesting that he is speaking to an audience. The man appears to be bald, and is looking slightly to his left, indicating that he is engaged in conversation or addressing someone or something off-camera. The background is a plain blue wall, which provides a simple and uncluttered backdrop for the speaker. The style of the video is straightforward and professional, focusing on the speaker and his message without any distracting elements."
+    ],
+    "video_ids": [
+        "MNxlZ1DSOaU_32_76to246"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man wearing a suit and tie with a microphone attached to his ear.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a suit and a striped shirt, with a microphone attached to his ear. Although he is not wearing a tie, the core description of a man in a suit with a microphone in his ear is largely fulfilled. The presence of a bow tie in some frames is likely an error or misinterpretation, but it does not contradict the primary condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a man wearing a suit and tie tie tie, which a microphone positioned to his ear. The the is not wearing a tie, the presence elements of a man in a suit with a microphone attached his ear is largely fulfilled. The presence of the microphone tie in the frames is acceptable a error or aalignmentation of but it does not significantly the main description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5c24975489fb642aff24cc8830dec1a75bacfa2b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c1ec2e0593972e4dbc4973153d7bd2520f793736576493e67e5fa959496080b6
+size 132489
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..8ef0b35ff347aed4a469ccce765c924db3d3397a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of intense emotion from a soccer player. The player, dressed in a white and blue striped jersey, is seen with his hand covering his mouth, a clear sign of shock or disbelief. The setting is a soccer stadium, with the blurred background suggesting a large crowd of spectators. The player's focused gaze and the intensity of the moment suggest that the game is at a critical point. The style of the video is realistic, capturing the raw emotion of the player in a real-life setting."
+    ],
+    "video_ids": [
+        "U1swfC0ZE3Q_9_0to103"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A soccer player dressed in a white and blue striped jersey, with his hand covering his mouth.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a white jersey with blue stripes, which matches the description. His hand is covering his mouth, and although the jersey has yellow accents, the core description of white and blue striped jersey with hand covering mouth is fulfilled. The presence of additional elements (like the logo and background) does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a white and with blue stripes, which matches the description. The hand is covering his mouth, which he the jersey appears additional stripes, it core description of the and blue stripes jersey is the covering the is fulfilled. The background of a elements likethe the crowd and the) does not contradict the main description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b9188a4494d5c89e56dba4a588252ca63964dd1f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b2476e86d448b46924a85db4ce3b43559611b7c5d04768f66e93cf3b418216d
+size 197868
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..0400b7915d60e29a7039a5ed3fdf7df368e17128
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures an underwater scene featuring a school of salmon swimming through a rocky riverbed. The water is clear enough to see the pebbles and stones at the bottom, which vary in size and color, ranging from light gray to dark brown. The salmon, with their distinctive reddish-orange bodies and silvery sides, swim in a coordinated manner, their fins and tails moving rhythmically as they navigate the current. Some fish are closer to the camera, while others are further away, creating a sense of depth. The water's surface is visible above them, with ripples and reflections indicating movement on the water's surface. The lighting is natural, suggesting the video was taken during daylight hours, and it highlights the vibrant colors of the salmon and the textures of the rocks below. The overall atmosphere is serene and natural, showcasing the beauty of these fish in their habitat."
+    ],
+    "video_ids": [
+        "936477b8d8c573905f09aadfac3366a74748ba4605a445c30fba9282379a712a"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A school of salmon with reddish-orange bodies and silvery sides.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple salmon swimming underwater, with their bodies displaying a reddish-orange hue and silvery sides, which matches the description. The underwater environment with rocks and clear water also supports the natural habitat of salmon. While some fish may appear more reddish or less distinct due to lighting or motion, the overall appearance aligns with the described characteristics.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a fish with in, which their bodies displaying a reddish-orange color and silvery sides, which align the description of The presence environment, colorful and water water also supports the setting habitat of salmon. The the additional appear appear slightly prominentish than sil sil, to the or angle, the overall depiction aligns with the given ' of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..faa62e49207af900876ff65f541998ef1a48283f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fcf74ff647b9c779d17f7b6672209120a472dec0e9f62efc5cd13c63cb288805
+size 327428
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..aaa0da265cdf4dae04f01b013f075079114394f0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a colorful and lively 3D animation featuring a large blue elephant and a group of small animals. The elephant, with its long trunk and tusks, is the central figure in the scene. It is standing on a sandy ground, surrounded by a lush green landscape with trees and bushes. The small animals, including a dog, a cat, and a pig, are gathered around the elephant, looking up at it with curiosity and admiration. The scene is set against a clear blue sky with a few clouds. The overall style of the video is playful and whimsical, with a focus on the interaction between the elephant and the small animals."
+    ],
+    "video_ids": [
+        "Dxm9-sCSAZo_252_0to174"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large blue elephant with a long trunk and tusks, a dog, a cat, and a pig.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a large blue elephant with a long trunk and tusks, along with a dog, a cat, and a pig. These core objects are present and accurately depicted, fulfilling the 'Object(s)' condition. Additional elements like a boy, a parrot, and other animals do not contradict the description and are acceptable as long as they don't conflict with the core objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a large blue elephant with a long trunk and tusks, which with a dog, a cat, and a pig. The elements elements match present and match depicted, fulfilling the 'Object(s)' condition.\"\n The elements like the tree and a treerot, and a animals are not contradict the core and are acceptable.\"\n long as they do't overshadow with the core elements.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..444ad4cce044455b78dc2fd44b4456b5ced9725c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb56a66f046d1a0d610f543c1ffa07dfe40b77da84d4a8209e8360eeb68d6169
+size 182191
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..86cd1529ffb84c724229e00ea96c9c2447cf228d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a mountainous landscape. In the first frame, a person is seen standing on a grassy hill, overlooking a winding river that cuts through the valley. The person is holding a camera, possibly capturing the breathtaking view. In the second frame, the person is seen walking down the hill, following the river's path. The third frame shows the person standing near the river, with the majestic mountains in the background. The video is shot from a high angle, providing a panoramic view of the landscape. The style of the video is naturalistic, capturing the beauty of the outdoors in a realistic manner."
+    ],
+    "video_ids": [
+        "DUPHXtSRPm8_49_0to116"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a grassy hill, a winding river, majestic mountains? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a person (a man and a woman) on a grassy hill, a winding river flowing through the valley, and majestic mountains partially covered in mist. All core elements from the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a person walkinga figure in a woman) walking a grassy hill, with winding river, through the landscape, and majestic mountains in covered with snow. The the elements of the description are present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ce81ad9ab7065a9c69c44e6a24f460d119325ecc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab1d540ec17dfe4b262dc7396f665ae73343c7e214fffc87e4a5d25c16b56fb7
+size 199093
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..6424c4d9ca10c1ae5eb6891c97d5e3ee73a93b2b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen in a kitchen setting, pouring a white substance from a container into a glass jar. The man is wearing a blue t-shirt with a colorful graphic on it. The kitchen is equipped with a refrigerator and an oven, and there are various items on the counter, including a bottle and a bowl. The man appears to be in the process of preparing a meal or a drink. The style of the video is casual and informal, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "WM_6dAY6liU_13_60to270"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue t-shirt with a colorful graphic, pouring a white substance from a container into a glass jar.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue t-shirt with a colorful graphic (a rainbow and cartoon characters) pouring a white substance from a container into a glass jar. The core action and subject match the description. Additional elements like the kitchen background, Homer Simpson sticker, and other items do not contradict the main description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a blue t-shirt with a colorful graphic,matching styl-like the-like). pouring a white substance from a container into a glass jar. The setting elements and the match the description provided The elements like the kitchen setting and other Simpson's, and other kitchen on not contradict the main action.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..10ea0685bc75475e33a1c728477bd19a45664baa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b13aa6b863dea537c1524c9d2b647c77bdd094d306301572f14c5c186e1d7445
+size 230013
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..9b0b60c6e69399c74db3bf769603df1e03aa3cce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a lively scene of a group of people gathered in a room with a blue background. The central figure is a man holding a microphone, who appears to be the host of a show. He is surrounded by a diverse group of individuals, all of whom are engaged in conversation and laughter. The atmosphere is one of camaraderie and enjoyment, with the participants actively participating in the discussion. The style of the video suggests a casual, informal setting, with the participants seated comfortably and the host standing to facilitate the conversation. The overall mood of the video is light-hearted and fun, with the participants clearly enjoying their time together."
+    ],
+    "video_ids": [
+        "Tl16jvOb4jk_122_36to187"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man holding a microphone (host), a diverse group of individuals engaged in conversation and laughter.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man holding a microphone, likely acting as a host, and a diverse group of individuals seated around him, engaged in conversation and laughter. The scene matches the described condition, with no clear contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man holding a microphone, which acting as a host, engaging a group group of individuals who around him. engaged in conversation and laughter. The setting suggests the description ' of with the additional contradictions or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8f18d0f9781511356caf268d152122438ab574eb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28add6addafb7186a6142c6dae575d577c9b16bb5295013958083dc4386c5fc0
+size 188609
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..5cb8809934ac3b1f651497d1ee7136a96703923c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features two shirtless men standing side by side, showcasing their tattoos and physiques. They are outdoors, with a backdrop of a vibrant sunset or sunrise, casting a warm glow on the scene. The men are posing confidently, with one pointing towards the camera, suggesting a sense of camaraderie and shared pride in their appearance. The overall style of the video is casual and relaxed, with a focus on the men's bodies and the natural beauty of the sky."
+    ],
+    "video_ids": [
+        "1qHSyHs_tf0_15_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two shirtless men posing side by side, one pointing towards the camera.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two shirtless men standing side by side, both with visible tattoos. The man on the right is pointing towards the camera with his right hand, matching the described action. The background and attire of the men are consistent with the description, and no elements contradict the core requirement.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two shirtless men standing side by side against and with visible tattoos. One man on the right is pointing towards the camera with his right hand, which the description action. The background features their of the men also consistent with the description, and there additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..278e3442bcc5ce4434baab75f7f06e2c31a2d926
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1229-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88b55e14e990b57c31c559ada896f1ce1c22c40ab8fb1f26f0abdc18feeb69ba
+size 153665
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..8c002de5b8600f62b06ee5e346864fa196253347
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a tender moment between a mother cat and her kittens in a grassy field. The mother cat, with her white fur, is seen licking the head of one of her kittens, which is black and gray in color. The kitten, with its eyes wide open, seems to be enjoying the affectionate gesture. The mother cat is also seen walking away from the camera, with her kittens following her. The scene is set against a backdrop of a grassy field, with a few trees visible in the distance. The video is shot in a realistic style, capturing the natural behavior of the cats in their outdoor environment."
+    ],
+    "video_ids": [
+        "aJOtGx5ue40_24_0to138"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Mother cat, kittens? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a mother cat grooming her kittens in a grassy outdoor setting, which matches the 'Object(s)' condition of mother cat and kittens. The presence of other animals in the background does not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a white cat and her kitten, a naturaly outdoor setting. which align the descriptionMother(s)' condition of the cat and kittens. The presence of the elements or the background does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a63bc66ad2b5587ea97fec25998fa68e6e7721f6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6787acc78265fe81dcc56196df74f6f02471c87abb9de74e606c5996d8c860db
+size 207749
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..3ba1071f392bdc48689b9978e695be090e36cf95
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video opens with an underwater scene inside a large aquarium. The water is a clear blue, and various fish swim gracefully through the frame. The fish are diverse in size and color, ranging from small, slender fish to larger, more robust species. In the background, a shark glides smoothly near the surface, adding a sense of depth and scale to the environment. The aquarium's structure includes several pillars and platforms, which serve as resting spots for the fish and provide a sense of scale. The lighting is soft and diffused, creating a serene and tranquil atmosphere.  As the video progresses, the perspective shifts slightly, offering a broader view of the aquarium's interior. The fish continue their graceful movements, and the shark remains a prominent feature in the background. The pillars and platforms become more visible, showcasing the architectural design of the aquarium. The overall ambiance remains calm and peaceful, emphasizing the beauty and tranquility of the underwater world.  The scene then transitions to a different setting, where two individuals are"
+    ],
+    "video_ids": [
+        "43de78d734345bd2610e1e9b1614dbdd48875c27fb99c9e4ba677216d5d93ec2"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Fish, a shark, pillars, platforms? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows various fish swimming in an aquarium environment, a shark visible near the top, and distinct pillars and platforms that appear to be part of the aquarium's structure. These elements align with the described 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful the elements swimming around an underwater-like, a shark gl in the bottom, and multiple pillars and platforms that resemble to be part of the aquarium's structure. The elements align with the ' 'Object(s)' condition.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cad3718b89db44ada0834797b5322224458edbe3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3cc8a9aa34eec5af045137c23fdcab1160b0ef6bb84a58d45f073e385105cfa2
+size 283772
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..8e6a8ed4a057074df14d37d3c24ddd4128fc4c8f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman with tattoos and a black turtleneck stands confidently in front of a body of water, with a tree and a rocky shore in the background. She is the central figure in the image, and her presence is commanding. To her left, a man with a beard and a floral shirt is seated, smiling at her. He appears relaxed and amused. To her right, another man with a mustache and a colorful jacket is seated, holding a book and looking at her with interest. The setting is serene and natural, with the water and the rocky shore providing a beautiful backdrop. The overall style of the video is casual and relaxed, with the three individuals appearing to be enjoying a leisurely day outdoors."
+    ],
+    "video_ids": [
+        "AE9UgQ74FSk_125_22to152"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with tattoos and a black turtleneck, a man with a beard and a floral shirt, another man with a mustache and a colorful jacket.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with visible tattoos and wearing a black turtleneck, seated between two men. One man has a beard and is wearing a floral shirt, while the other has a mustache and a colorful jacket. All three individuals match the description provided, and no elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a woman wearing tattoos tattoos and wearing a black turtleneck, a next two men. One man has a beard and is wearing a floral shirt, while the other man a mustache and is colorful jacket. The elements individuals are the description provided, indicating the additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4ec13851763f76d327191e29d8d8a731f18a941a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ad0c1d5b971093e27ee0f1467126afd39a9310f0de15b43724113b475872a392
+size 248826
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..d8b014fd89cacab00ddaf899bbbad8713b7183b3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features two animated cars, one orange and one green, driving down a street lined with trees and a wooden fence. The orange car has a large, expressive face with eyes and a mouth, while the green car has a smaller face with eyes and a nose. Both cars have round headlights and a grill that resembles a mouth. The orange car has a large grill and a bumper with a license plate, while the green car has a smaller grill and a bumper with a license plate. The cars are driving on a paved road, and the trees and fence are in the background. The video is in a cartoon style with bright colors and exaggerated features."
+    ],
+    "video_ids": [
+        "Ppu-CssXBmE_13_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two animated cars (one orange and one green), both with faces, headlights, and grill resembling a mouth. The orange car has a large grill and bumper with a license plate, while the green car has a smaller grill and bumper with a license plate.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features two animated cars with faces, headlights, and grills that resemble mouths, matching the description. The orange car has a large grill and bumper, while the green car has a smaller grill and bumper. Both cars also appear to have license plates, even though they are not clearly visible in detail. The presence of additional elements like the hamburger on the green car's roof does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features two animated cars, faces, headlights, and aills resembling resemble mouths. matching the description. The orange car has a large grill and bumper with and the green car has a smaller grill and bumper, Both cars have have to have license plates, which though the are not clearly visible in the. The presence of a elements like the road and the orange car's roof and not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..97b6309312f2793deaa2328357de3d73d5c74123
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b10c614906ed3be86a2b59a6d13c426eda14a2df738ea53300856b0da6675ad
+size 293087
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..0d41c95d18a66c895f8721d6fe02c761a2aba99a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are standing next to a red car with the trunk open. The man is gesturing towards the car, possibly explaining something to the woman. The car is parked on a street, and there are trees and a house in the background. The woman is wearing a black hoodie, and the man is wearing a blue jacket. The video captures a casual, everyday moment between the two individuals."
+    ],
+    "video_ids": [
+        "VbCmXNud6FA_65_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman, a red car with an open trunk? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman standing next to a red car with its trunk open, which matches the core description. Additional elements like a spare tire and the website watermark do not contradict the primary condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man and a woman standing next to a red car with its trunk open. which matches the description description of The elements such the grass tire and a grass logo do not contradict the main condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0e427d59767c70fb9ad83a8f333e59497c7cf4b7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5701dd5894e864083372fdec03f542238c58e398dda28b18ce2ae5983faf159
+size 230825
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..c87bf511b118178d0a9c9dbba5e6a511f6a2fca9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene day at a college campus. The camera follows a group of students as they walk along a path lined with trees and buildings. The students are casually dressed, some carrying backpacks, suggesting a typical day of classes or activities. The path leads towards a large, imposing building with a tall tower, which stands out against the clear blue sky. The building's architecture is grand and impressive, hinting at the institution's historical significance. The overall atmosphere of the video is peaceful and academic, with the students and the campus serving as the main subjects."
+    ],
+    "video_ids": [
+        "-935PQmP_08_189_0to162"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A group of casually dressed students walking along the path, some carrying backpacks.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows several individuals walking along a path in what appears to be a university campus. They are dressed casually, and some are carrying backpacks, which aligns with the description of 'a group of casually dressed students walking along the path, some carrying backpacks.' The presence of the iconic tower and bare trees does not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a individuals walking along a path, a appears to be a campus or setting They are dressed casually, and some are carrying backpacks, which aligns with the description of studentsA group of casually dressed students walking along the path, some carrying backpacks.' The setting of the university building and the trees suggests not contradict the description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ab8f18c075e843005b3b6bdea26b3b46e62d27d6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c10c73f34c6fe9d087e58f30cd7567f124ef038c84851f32a7ca520b1fd044d
+size 266831
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..fed3feff5145b06f92663b140bc94037e01d8197
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a black Ram truck parked on a rocky desert terrain during a dramatic sunset. The truck is positioned in the center of the frame, with its headlights on, illuminating the surrounding area. The sky transitions from a warm orange to a deep purple as the sun sets, casting a soft glow on the landscape. In the background, a mountain range can be seen, adding to the sense of vastness and isolation. The overall style of the video is cinematic, with a focus on the interplay between the natural beauty of the desert and the ruggedness of the Ram truck."
+    ],
+    "video_ids": [
+        "cIkBTeceRvQ_16_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black Ram truck, a mountain range? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a black Ram truck prominently in the foreground, and a mountain range is visible in the background under a vibrant sunset sky. These elements align with the specified 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a black Ram truck prominently in the foreground, which a mountain range is visible in the background during a sunset sunset sky. The elements align with the description 'Object(s)' condition,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..949bd26fb4e3f9fce2897e3288e188155d885667
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6416a7edaa9ce2a3299092ea6c4e7045558fc6da4e83162b30782f51deb56a67
+size 143223
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..464578ea7d3e7148a1661cacb3d97518743004d9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a silver SUV parked on a sandy beach. The car is positioned facing the ocean, with the waves gently lapping at the shore. The sky is clear and blue, suggesting a sunny day. In the background, there are mountains visible, adding to the scenic beauty of the location. The car appears to be stationary, with no people or other vehicles in sight. The overall style of the video is serene and peaceful, capturing the tranquility of the beach setting."
+    ],
+    "video_ids": [
+        "U-87WfT3SRw_27_0to134"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A silver SUV parked facing the ocean with waves gently lapping at the shore.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a silver SUV parked on a sandy beach, with the ocean and waves visible in the background. The SUV is positioned facing the ocean, and waves can be seen gently lapping at the shore, matching the description. The presence of hills and sky does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a silver SUV parked on a sandy beach with facing the ocean and waves visible in the background. The car is positioned facing the ocean, and the can be seen gently lapping at the shore. matching the description provided The presence of a in a in not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e787072c73480d948c84e1643c12be3e37c42db0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-123-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae204159d2907bd014c8e613efc342e184f032f7b71ec96817e50987a47ddfc3
+size 170934
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..7424576d391e8644246a17cfb7fcd0fef441fe72
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a delicious meal consisting of a burger and fries. The burger, with its golden bun and a variety of toppings, is the main focus of the video. It's placed on a piece of aluminum foil, which is resting on a brown paper bag. The fries, golden and crispy, are scattered around the burger, adding to the appetizing scene. The entire meal is presented in a way that makes it look very tasty and inviting. The style of the video is simple yet effective, focusing on the food and making it the star of the show."
+    ],
+    "video_ids": [
+        "p4VCQcXPAv0_8_0to122"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Burger (golden bun, various toppings), fries (golden, crispy)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a burger with a golden bun topped with sesame seeds and various toppings such as lettuce, cheese, and bacon. It also displays golden, crispy fries. These elements match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a burger with a golden bun and with various seeds, various ingredients such as lettuce, tomato, and what. Additionally is features a, crispy fries. The elements align the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ab1feefb9e6c156f645f3a9e25b658b72e7d566a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f0643ad3e8c0981d1eec4ce95b020d3546482aa9eaefc59f87e91d214a357864
+size 144842
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..5336a12d5d359e1b34dc3232b7c30c2fc7f7ba0a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated sequence featuring three vehicles and a dog. In the first frame, a red fire truck with a ladder on top is parked on the left side of the road. The fire truck has a sad expression on its face. In the second frame, a white ambulance with red stripes is parked on the right side of the road. The ambulance has a happy expression on its face. In the third frame, a small brown dog is standing in the middle of the road between the fire truck and the ambulance. The dog is looking at the camera with a curious expression. The background of the video is a gray road with a blue sky and white clouds."
+    ],
+    "video_ids": [
+        "bMAcdiL0rTQ_69_0to119"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red fire truck, a white ambulance, and a small brown dog? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red fire truck, a white ambulance, and a small brown dog, all of which match the described objects. The background and additional elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a red fire truck, a white ambulance, and a small brown dog. which of which match the description objects. The presence and setting elements do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..90e4cf9b822103046d1b6ecef6c772452fb66694
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:65a3d4a809dfb2d2164d88e33af42cbc72090024168070c179d0bfee55c58539
+size 164672
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..02d828738eaf0ec0ba6a14a7807bb149c9dc7ad4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a blue baseball cap and a blue jacket, standing on a field. He is looking to the side with a serious expression on his face. The background is blurred, but it appears to be a sports stadium with a crowd of spectators. The man's posture and expression suggest that he is focused on something happening off-camera. The style of the video is a standard sports documentary, with a focus on the athlete's expression and the atmosphere of the stadium."
+    ],
+    "video_ids": [
+        "I0__0eHxafM_62_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue baseball cap and a blue jacket, standing on a field, looking to the side with a serious expression.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue baseball cap with a logo and a blue jacket, standing on what appears to be a field, looking to the side with a serious expression. Another person is partially visible in the background, but the main subject matches the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue baseball cap and a logo and a blue jacket, standing on a appears to be a field. looking to the side with a serious expression. The person is partially visible in the background, but this main subject matches the description provided}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2cb98ff722a865d12f0c02198c4c83aa145767ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4227558f3f1476abd99daba835a2096fb09b344accccda4af66c661ebffdd6e9
+size 168156
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..0557fd39ff2c939389470be6dca7fd8200ba2154
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a black sports car parked on a paved surface, possibly a parking lot or a street. The car has a sleek design with a low and wide stance, featuring large black wheels with orange accents. The car's body is shiny, reflecting the surrounding environment. In the background, there are other vehicles parked, including a white van and a red car. The setting appears to be an urban or suburban area with a building and a fence visible in the distance. The style of the video is straightforward, with no visible text or additional graphics. The focus is on the car, and the camera angle is slightly elevated, providing a clear view of the vehicle's design and color. The lighting suggests it might be daytime, and the overall atmosphere is calm and quiet."
+    ],
+    "video_ids": [
+        "dPr7LOjIKVU_31_0to146"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Black sports car, white van, red car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a black sports car, which matches the core description. In the background, a white van and a red car are also visible, fulfilling the additional elements mentioned in the condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a black sports car, which is the description description. Additionally the background, there white van and a red car are also visible, fulfilling the ' elements of in the condition. conflicting.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4b23aa7229d0058550963c7b0cc3313ed51a7271
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a216b2c7a921bba6bdf8f936edad5cf2ca4d4e827407f6f389bf8adef142feef
+size 169958
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..d8cb76868e882e3b82a9732a91893f1d7265fd0c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic scene on a race track. A red car is seen speeding down the track, its vibrant color contrasting with the gray asphalt. The track is marked with black and yellow stripes, adding a sense of speed and motion to the scene. The car is positioned in the center of the frame, drawing the viewer's attention to its movement. In the background, a building can be seen, providing context to the location of the race track. The overall style of the video is dynamic and action-packed, capturing the thrill of racing."
+    ],
+    "video_ids": [
+        "yD6ExK78IXc_24_0to169"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video consistently features a red car as the primary subject throughout all frames. The car is shown from various angles and perspectives, including side views, rear views, and aerial shots, all of which clearly depict a red car on a racetrack. There are no elements in the video that contradict the presence of a red car.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a red car driving the main object, the frames. The car is depicted driving a angles as positions as maintaining a views and as views, and close views, all of which confirm depict the red vehicle. a racetrack. The are no elements in the video that contradict the presence of a red car.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6174038204ba8aeb528eaa1bd36475034592e3d4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:228f0581d534ff0fc5072ddde571ae850306ae8042e9bad38768e3588746e172
+size 228007
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..6042dd91979a9fbcaa33724a5540debb9a775bbe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a black shirt with a white collar, sitting in front of a blue background. He appears to be speaking or reacting to something, as his mouth is open and his eyes are closed. The man is the main subject of the video, and there are no other objects or people visible. The style of the video is a close-up shot, focusing on the man's face and upper body. The lighting is bright, highlighting the man's features and the blue background. The video does not contain any text or additional graphics. The overall impression is that the man is engaged in a conversation or reacting to a situation, but the specific content of his speech or the context of the situation is not clear from the image alone."
+    ],
+    "video_ids": [
+        "KN31FtPgcfE_4_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a black shirt with a white collar? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black shirt with a visible white collar underneath. This matches the description exactly, and no conflicting elements are present that would contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black shirt with a white white collar.. The matches the description provided, indicating there additional elements are present.\"\n would contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a59c60ce20b1bc66b12499458d04f3e4dfb35693
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37cfbf205df4380002922a2f19a14f777f3b5e36e3574142a7a18b47ec9a27c6
+size 144586
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..73dfb0330954b7b2469c8eeca6706f90440d7843
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car, specifically a Ford Ranger, from the perspective of the driver's seat. The car is parked in a showroom, as evidenced by the other cars visible in the background. The interior is well-lit, with the dashboard and center console illuminated by the car's lights. The dashboard features a touch screen display, and the center console houses the gear shift and climate control system. The car's seats are upholstered in a light-colored material, and the door panels are also light-colored, with the door handles and window controls visible. The car's steering wheel is on the left side, indicating that it is designed for driving on the right side of the road. The overall style of the video is straightforward and informative, providing a clear view of the car's interior features."
+    ],
+    "video_ids": [
+        "fKcenYxRxPM_35_78to231"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Ford Ranger (car), dashboard, center console, touch screen display, gear shift, climate control system, seats, door panels, steering wheel? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the interior of a Ford Ranger, including the dashboard, center console, touch screen display, gear shift, climate control system, seats, door panels, and steering wheel. All specified elements are visible and accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts the interior of a vehicle Ranger, including the dashboard, center console, touch screen display, gear shift, climate control system, seats, door panels, and steering wheel. The these elements are present and match depicted, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e668583ce37f14dc5717d41b115b9d4c61c54a64
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:55626f2159a5a26847df9d2d18d34c485d8df38653aafc59c27f8fb78fda39b1
+size 72783
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..1bcb4c32deeb2bcc0dda5fec2d565e741969b1f5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases the interior of a luxury car, focusing on the center console and the gear shift. The car's interior is predominantly white, with red accents adding a touch of elegance. The gear shift, which is the main focus of the video, is made of chrome and has a red and silver color scheme. The video captures the gear shift in three different positions: neutral, drive, and reverse. The gear shift is shown in detail, highlighting its design and functionality. The overall style of the video is sleek and sophisticated, reflecting the luxury and refinement of the car's interior."
+    ],
+    "video_ids": [
+        "8d0c5Ps_BYI_21_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Center console, gear shift? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the center console and gear shift of a luxury car interior. The gear shift is prominently displayed, and the surrounding center console features chrome accents and controls, matching the description. The camera movement provides a detailed view of these elements without contradicting the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a center console of gear shift of a car vehicle.. The gear shift is prominently displayed in and the center area console, various accents and buttons, which the description of There presence angle and a clear view of these elements, anying the ' description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..545d2e1b0f283c65f0ab064f37753e72eb11ec05
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1230-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:924fb1afe6d285b5c2efc4bdaade82b98d5cf38e0d63f333b9c97bb433eeb1c4
+size 70508
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..251b15c95be8d646d2704e6794bc944bde5099bc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a culinary journey featuring a bowl of tomato soup. The first frame shows a close-up of the soup, with a spoonful of butter being added to the bowl. The second frame captures the butter melting into the soup, creating a rich and creamy texture. The third frame reveals the final product, a bowl of delicious tomato soup with a dollop of butter on top. The style of the video is simple yet effective, focusing on the preparation and presentation of the soup. The use of close-ups and slow motion shots highlights the details of the ingredients and the cooking process, making the viewer feel as if they are in the kitchen with the chef. The overall effect is a mouth-watering visual feast that showcases the beauty of simple, home-cooked meals."
+    ],
+    "video_ids": [
+        "vzKr78XucbE_10_0to155"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bowl of tomato soup, spoonful of butter? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a bowl of tomato soup with a spoonful of butter on top, which matches the description. The surrounding tomatoes and bamboo mat are additional elements that do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a bowl of tomato soup in a spoonful of butter being top. which matches the description of The sequence elements and green mat in additional elements that do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dfa7151b68e531773f944306c6b67b2f68406a20
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fdc85c0713fb0dc712e70bf3554f5005dbbc5d790c9696600e960a5a419c8409
+size 129179
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..0645840ad429d78a5c5f5987e1cc4c317ed4e07c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with red hair and a green top, smiling and looking to her right. She is seated in front of a Christmas tree, which is adorned with lights and ornaments. The woman appears to be engaged in a conversation with another woman, who is not visible in the frame. The video has a festive and cheerful atmosphere, with the Christmas tree serving as a prominent background element. The woman's smile and the overall setting suggest a warm and joyful holiday season."
+    ],
+    "video_ids": [
+        "EP5HU0UlbjY_20_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with red hair and a green top, smiling and looking to her right. Another woman (not visible) with whom she is having a conversation.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with red hair and a green top, smiling and looking to her right, which matches the description. Although another woman is partially visible to her right, she is not fully in frame, making her 'not visible' as described. The core description is largely fulfilled without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with red hair and a green top, who and looking to her right. which align the description. The the woman is partially visible and the right, the is not fully visible the, and it presencenot visible' as per. The setting elements is largely fulfilled.\"\n contradict.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1746c636b28fd9b6b9be5051d95d47258c185052
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7263ae48622c8975ad2ccff1588d1fc40409f0d2bdfbfc383ec3aed109558cc5
+size 162017
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..1a6bc140b5491ab1f1bdbd1dbde98fba018ac8a8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a heart and a brain, anthropomorphized as characters, walking together in a white, reflective environment. The heart, colored in a vibrant red, is on the left side of the frame, while the brain, depicted in a realistic white, is on the right. They appear to be holding hands, symbolizing a connection between the two. The heart is slightly ahead of the brain, suggesting a lead or guidance role. The overall style of the video is simplistic and clean, with a focus on the two characters and their interaction. The white background enhances the visibility of the characters and their actions. The video does not contain any additional objects or elements, keeping the focus solely on the heart and brain characters."
+    ],
+    "video_ids": [
+        "RCa6ULGHsa8_34_0to198"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Heart and brain, both anthropomorphized as characters.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by depicting a heart and a brain, both anthropomorphized as characters, holding hands and walking. The heart is red and has limbs, and the brain is gray with limbs, clearly anthropomorphized. The transition to a coconut oil scene does not contradict this core description, as it appears to be a separate scene or transition, not undermining the initial anthropomorphized characters.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features fulfills the 'Object(s)' condition as depicting a heart and a brain as both ofomorphized as characters. with hands. walking together The heart is red and heart a, while the brain is white with a and indicating showingomorphized. The interaction between a more character bottle does not affect the description description as as it is to be a separate scene and a in not part the anthrop depictionomorphized characters.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a7449269a968b740facc1e301a90744af554f0d3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f8450b9aa80148e3508f73570c4d948dc746971d4b6070c86c719794e58c044d
+size 138385
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..7528ca94c730398e4ecb3863f83dbdc466844239
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a man in a white shirt and blue tie. He is seated in front of a blurred background that suggests an indoor setting, possibly a news studio. The man appears to be speaking or reacting to a question, as indicated by his facial expression and the slight movement of his mouth. The style of the video is typical of a news broadcast, with a focus on the interviewee and a professional, polished appearance. The lighting is bright and even, highlighting the man's features and the details of his attire. The overall impression is one of seriousness and professionalism, consistent with the expectations of a news program."
+    ],
+    "video_ids": [
+        "9Z-YQZqdRLM_17_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a white shirt and blue tie, seated.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a white shirt and blue tie, seated in what appears to be a studio setting. The core description is accurately represented, with no significant contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a white shirt and a tie, seated in what appears to be a studio setting. The description description of met represented in with no additional contradictions or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3452b4a62861c5420fa20aff5b9c956d502286d5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4e67d8504d10746c1beedce523f98246ea0c700eb290d84cced112d70f6f1af
+size 155138
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..93b94118662197a4caec571a24f7169ab0f06818
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red sports car parked in a garage. In the first frame, the car door is open, revealing the interior. In the second frame, a man is seen taking a photograph of the car. In the third frame, the man is still taking the photograph, but the car door is now closed. The car is the main focus of the video, and the garage setting provides a contrasting backdrop to the vibrant red of the car. The man's action of taking a photograph adds a dynamic element to the otherwise static scene. The video captures the beauty of the car and the garage setting, creating a visually appealing composition."
+    ],
+    "video_ids": [
+        "JN6dxWVrTKM_9_0to145"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red sports car, a man? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red sports car with its door open and a man leaning on the rear of the car. The car is prominently featured, and the man is visible interacting with the vehicle. Additional elements such as other people and vehicles in the background do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a red sports car with a doors open, a man interacting into the car of the car. The car's the featured, and the man is interacting, with the vehicle, The elements such as the cars and the in the background do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3224b2f60168e26aae4638096dd3ee0ecc8d24ea
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9b581fdbb681b5567b4a1700fb4dfd030945de1e16c9791d6557d35e864ae11f
+size 199968
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..eb9c30e52c60a61b75e9482b4dc998928ce25286
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated scene featuring two characters, a young boy and a girl, in a room with a pink curtain. The boy is holding a yellow star and is smiling at the girl. The girl is looking at the star with a curious expression. The room has a colorful and playful atmosphere. The animation style is cartoonish and child-friendly."
+    ],
+    "video_ids": [
+        "2nI5jOCqLbY_135_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two characters (a young boy and a girl), a yellow star.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features two characters: a young boy and a girl, as described. The boy is holding a yellow star, and the girl interacts with the baby and the star. The core elements of the description are present without contradiction, even though additional background elements like windows and furniture are visible.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts two characters, a young boy and a girl, and described. The girl is holding a yellow star, which the girl is with him star, the star. The setting elements of the description are present, any.\"\n and though the elements elements like the and curtains are visible,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3287540242e62face88d33b51d966014c5da5eab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c5d6b67ecea10f44a77befd0a210e4faef1589704db2680054bf82f0e58c3be1
+size 166014
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..833d5f732ec9a3eecb35966bbadd2ad50969a775
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man wearing sunglasses and a blue shirt, smiling and looking to the side. The style of the video is casual and friendly, with a focus on the man's facial expression and the way he interacts with his surroundings. The background is simple and uncluttered, allowing the viewer to focus on the man and his expression. The lighting is soft and natural, suggesting an outdoor setting. The overall mood of the video is positive and relaxed."
+    ],
+    "video_ids": [
+        "APMD0pI1fyM_10_0to193"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man wearing sunglasses and a blue shirt, smiling and looking to the side.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing sunglasses and a blue shirt, and he is smiling while looking to the side. These elements match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing sunglasses and a blue shirt, and he is smiling while looking to the side. These elements match the description provided, the questionObject(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e087c57f0e3599adce093c00c42991c474bca187
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:599c4ba094c0b0d14bfeff9949d6d4838d072912946b923ebf729dfebab62a1f
+size 211628
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..4261f43e2091d8c8279fc0c86d70349a093ee44e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person in a red apron preparing a meal in a kitchen. The person is using a silver pot on a gas stove, and a black and silver pressure cooker is nearby. The pressure cooker has a digital display showing the time as 03:45. The person is pouring something from the pot into the pressure cooker. The kitchen has a modern design with a white countertop and a black stove. The person is focused on the task at hand, and the overall atmosphere of the video is one of careful meal preparation."
+    ],
+    "video_ids": [
+        "Yd-SNqOf0-U_11_0to102"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person in a red apron, silver pot, gas stove, black and silver pressure cooker, white countertop, black stove.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition. It clearly shows a person wearing a red apron, a silver pot being placed on a gas stove, a black and silver pressure cooker (with visible brand 'Instant Pot' and digital display), and a white countertop beneath the stove. The stove itself is black, matching the description. All core elements are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as It shows shows a person wearing a red apron, pouring silver pot, used on a gas stove, and black and silver pressure cooker,which a brand 'T Pot'), on ' display), and a white countertop. the stove. The presence is is black, and the description. The elements elements of present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a015082de83430cc4960720e6dc482cc8f8925a0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1231-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d8d7badaf56967f0d959843cdadbb97f6c99a5e61967c9e270096225a3dec4a4
+size 143500
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..630382373e28c798619df033916aaf86ae65bbdb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a pond teeming with life. The surface of the water is adorned with numerous lily pads, varying in shades from vibrant green to hints of yellow and brown, suggesting a mix of fresh and slightly aged leaves. Among these lily pads, two koi fish gracefully swim. One koi is predominantly white with black markings, while the other displays a striking pattern of red and white. Their movements create gentle ripples on the water's surface, adding a dynamic element to the tranquil setting. The background reveals more aquatic plants, contributing to the lush environment. The camera remains stationary throughout the sequence, allowing viewers to fully absorb the peaceful ambiance of the pond and the graceful dance of the koi fish amidst the floating lily pads."
+    ],
+    "video_ids": [
+        "4fd4621c5cf4a9bedb1e5a992dc2c601f78ea15cac027c8f8dd7bdc8d09f5f47"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two koi fish, one predominantly white with black markings, and another with a striking pattern of red and white.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two koi fish: one predominantly white with black markings and another with a striking red and white pattern. Both fish are visible swimming in the pond among lily pads, matching the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts two koi fish. one with white with black markings, another with a striking pattern and white pattern. The fish are swimming swimming among the water, theily pads, which the description provided.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6a65ee3c56a031b3c5d1f91ff3bad8a0bad8dd26
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a548757fe0cdf25a21a48d4a9877ad14d90ee8072d0322ff64a4c05c9c80bcb
+size 237027
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..32ea634b04d88dec43fc67894f305ff8334e85c3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a hockey game. The main focus is a hockey player, dressed in a white jersey with red and black stripes, who is in the midst of a powerful swing with his hockey stick. The player's mouth is open, suggesting a yell or a scream, indicating the intensity of the game. The player is on the ice, which is a typical setting for a hockey game. In the background, there are other players, some of whom are wearing orange jerseys, suggesting they are part of the opposing team. The presence of a crowd in the stands indicates that this is a professional game, likely taking place in a large stadium. The overall style of the video is action-packed and energetic, capturing the excitement and intensity of the sport."
+    ],
+    "video_ids": [
+        "2TZLpbLWEDs_46_0to140"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A hockey player in a white jersey with red and black stripes, other players in orange jerseys, and a crowd in the stands.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a hockey player in a white jersey with red and black stripes, which matches the description. Although the background crowd is visible, there are no other players in orange jerseys clearly visible in the frame. The presence of the crowd and advertising boards does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a hockey player in a white jersey with red and black stripes, which matches the description. There there player is is not, the are no other players in orange jerseys in visible in the video. The focus of the crowd in the boards in not contradict the description description but}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3c0c98f66714128ac3eb2ae33b2758057d12fe9e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ecb8a07d1708dbd0d45ff75d66b0333b6237677a01fa0e83eaf79d8dc7850f78
+size 234401
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..d92a74330f13981b70817b62da3db56d3d66c86a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a garage filled with various cars, each frame capturing a different perspective of the scene. The garage is spacious, with cars parked on the ground and on lifts. The cars are of different makes and models, and they are parked in a somewhat haphazard manner. The lighting in the garage is bright, illuminating the cars and casting shadows on the floor. The garage appears to be well-maintained, with a clean and organized appearance. The video is likely shot in a documentary style, capturing the everyday life of a car enthusiast or a mechanic."
+    ],
+    "video_ids": [
+        "_Xs_aDvOkdg_67_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Various cars of different makes and models.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video displays multiple cars, primarily BMW models, in various colors and styles, including yellow, orange, silver, and black. They are arranged on multi-level racks and parked on the floor, showing different makes and models, which fulfills the condition of 'Various cars of different makes and models.'\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a cars of including seds, which a colors such states, which a, blue, and, and blue. This are arranged in a-level parking, in in a ground, which a makes and models. which fulfills the ' of havingVarious cars of different makes and models.'\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cb21f3f003303ace3a09bc28852824c32c7fee48
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d1814be2a115b71030af2d6b4c2f2d5d6dd573ed8d55172951002af691b10f3
+size 163986
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..b707cc8f794b5d0e847f2eb5a6adb6c699dd8eaa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the majestic Stonehenge, a prehistoric monument located in Wiltshire, England. The three frames show the monument from different angles, revealing its unique circular arrangement of standing stones. The stones, weathered by time, stand tall against the backdrop of a clear blue sky. The lush green grass surrounding the monument adds a touch of nature to the scene. The video is a testament to the architectural prowess of the ancient people who built this monument, and it offers a glimpse into the rich history of the region."
+    ],
+    "video_ids": [
+        "0U97Zkb7PRY_28_243to434"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Standing stones? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Standing stones' condition by prominently featuring large, upright megalithic stones arranged in a circular formation, which is characteristic of standing stones. The scene is clearly focused on these structures, with no elements contradicting their presence.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object stones' condition as depicting featuring large, upright stonegalithic structures that in a circular formation, which is characteristic of Stone stones. The presence includes consistent set on these structures, and no conflicting thating the presence.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..61831ddc85badd5d17f96d310bfe191626cdb71d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5c05d67520783b24477973227eef77ce8d139c53d8be4fa33dfd064615d70ac
+size 213227
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..6ce89451b78d69db05985f3e6c0657056bdf5a96
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a soccer match. The main focus is a player in a red and white striped jersey, who is skillfully maneuvering the ball with his right foot. He is in the process of dribbling past a defender in a white jersey, who is attempting to block his path. The player's body language suggests he is in control of the ball and is looking to make a pass or shoot. The background reveals a packed stadium, with spectators eagerly watching the action unfold. The atmosphere is intense, with the crowd's anticipation palpable. The style of the video is a fast-paced, action-packed sports sequence, capturing the excitement and skill of the players on the field."
+    ],
+    "video_ids": [
+        "fz--QJM6wKM_36_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a red and white striped jersey, a defender in a white jersey, spectators in the stadium.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a player in a red and white striped jersey (Atletico Madrid) being challenged by a defender in a white jersey (Real Madrid). In the background, spectators are visible in the stadium, fulfilling all elements of the description. The scene is consistent with a soccer match, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a player in a red and white striped jersey,likelyletico Madrid) and pursued by a defender in a white jersey (Real Madrid). The the background, there are visible in the stadium, adding the the of the '. The presence is a with a soccer match, and the additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a9f26e9f9201c76357f14e10150f131834707dce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6055e6c98341c5b5d01a71fffc78ac3473b7e82fdd4f0517d6a96218a64f638b
+size 332016
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..929dc7c8436cbc90c42774fddb8a0f5487533f40
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up of a red car's rear end, showcasing its design and features. The car's taillights are prominently displayed, with a sleek and modern design. The car's body is shiny and well-maintained, reflecting the surrounding environment. The car's logo is visible, indicating its make and model. The video is likely a promotional or review video, highlighting the car's design and features. The style of the video is straightforward and focused, with a clear emphasis on the car's rear end and design elements."
+    ],
+    "video_ids": [
+        "Y_3xx_yCp4w_12_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red car's rear end, taillights, car's logo? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by prominently displaying the red car's rear end, including its taillights and the visible 'I-PACE' logo. The focus remains on these specified elements throughout the clip, with no conflicting or contradictory visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition by showcasing featuring the rear car's rear end, the the taillights and the car carGS5ACE' logo. The focus is on these elements elements, the frames, with no additional additional additional elements information.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..10c4faa6d1c2466137beea4f22b1e1c2abbdd326
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b6992a7e3995f025743630dc7c6fdf0793f55644349881cf5ad4ac7ddab5a0c
+size 154064
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..81edf3ace4079eb8f83af5c4a8a71bd69e30ad2a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a muffin baking process. The muffins are placed in green paper cups on a wire cooling rack. The muffins are golden brown and have a slightly crispy texture on top. The muffins are arranged in a grid pattern on the cooling rack. The wire cooling rack is placed on a dark surface. The muffins are baked to perfection and are ready to be served. The video captures the transformation of the muffins from uncooked to cooked, highlighting the baking process."
+    ],
+    "video_ids": [
+        "7fw_WqjbloM_5_408to624"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Muffins in green paper cups? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows freshly baked muffins in bright green paper cups arranged on a cooling rack. The muffins have a golden-brown top, and the green cups are consistently visible throughout the frames, fulfilling the 'Object(s): Muffins in green paper cups' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows muff baked muffins placed green green paper cups. on a baking rack. The descriptionins are a golden-brown topping, indicating the green cups are a visible, the frames, fulfilling the 'Object(s)' Muffins in green paper cups' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f1cba67cf328a9dd08790bca21762b8851b6d9d0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af6ac61058a3c7811527e5ffd8e77f80594512eccf1dc8c2d957317b144ed41a
+size 145101
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..1b8bcb2c137b3a638fc517312b6d584af633275e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of triumph on a football field. A player, wearing a white jersey with the number 87, is in the foreground, his mouth open in a triumphant yell. He is surrounded by his teammates, also in white jerseys, who are running towards him in celebration. The crowd in the background is a blur of red and white, indicating a large turnout for the game. The atmosphere is one of excitement and victory."
+    ],
+    "video_ids": [
+        "X6Ec-hZaToA_12_30to164"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a white jersey with number 87, his teammates in white jerseys, and a crowd.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by prominently featuring a player in a white jersey with number 87, along with his teammates in white jerseys, and a crowd in the background. The presence of the NFL Draft 2021 logo does not contradict the core description, as it is an additional element that does not conflict with the main subjects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting featuring a player in a white jersey with the 87, his with his teammates in white jerseys. and a crowd in the background. The players of the crowd logo logo2023 logo in not contradict the core description and as it is a additional element that does not conflict with the main focus and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..173f03c0068e02922456fc6f3937a977e8df5226
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1232-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57c5ed91d2e81341e4ad96275139b314183847501e7cbb26d64389cd5e8b986b
+size 312876
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..6479d45ffb6a6ab6d470c1c192995d132e0f904e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a football player in action, wearing a black jersey with the number 10 and neon green accents. He is holding a football in his right hand, ready to throw it. The player is wearing a black helmet with a visor, and his body language suggests he is focused and determined. The background is blurred, but it appears to be a football field with a crowd of spectators in the stands. The style of the video is dynamic and action-packed, capturing the intensity of the sport."
+    ],
+    "video_ids": [
+        "ZsfkaST6VQI_24_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A football player, a football, a black helmet with a visor, a black jersey with number 10 and neon green accents.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a black helmet with a visor, a black jersey with the number 10 and neon green accents, and holding a football. All core elements described in the condition are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a football player wearing a black helmet with a visor, a black jersey with the number 10 and neon green accents. and holding a football. The the elements of in the condition are present in accurately depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a4535df29ac89c47658a38296aa94808a6a4641a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e85d10fd0a2e48ab96e213597c4e33fc7b8eae6bb8157cdfda05357abcd0b63c
+size 202885
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..9a0cb6e344e05b2125cd690430de137a96a309f0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up of a red car's rear end, showcasing its design and features. The car's taillights are prominently displayed, with a sleek and modern design. The car's body is shiny and well-maintained, reflecting the surrounding environment. The car's logo is visible, indicating its make and model. The video is likely a promotional or review video, highlighting the car's design and features. The style of the video is straightforward and focused, with a clear emphasis on the car's rear end and design elements."
+    ],
+    "video_ids": [
+        "Y_3xx_yCp4w_12_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red car's rear end, taillights, car's logo? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by prominently displaying the red car's rear end, including its taillights and the visible 'I-PACE' logo. The focus remains on these specified elements throughout the clip, with no conflicting or contradictory visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition by showcasing featuring the rear car's rear end, the the taillights and the car carGTYZACE' logo. The image is on these elements elements, the frames, with no conflicting additional additional elements information.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b9412de90d4ab0204061779fd445d9be3a7cc680
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:715e8ac99cb60fcad4e09c48ac72eb8181766d9d172d0fe2767341c207814924
+size 135647
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..a7e3175d7226515e70e2fd56230dd4c78ab873d9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is a football player, wearing a blue and green uniform with the number 72 prominently displayed. He is in the midst of action, holding a football securely in his right hand. His helmet, matching his uniform, is equipped with a visor, adding to the intensity of the scene. The player's body language suggests he is in motion, possibly running or preparing to make a play. The background is a blur of colors, indicating the presence of other players and the crowd, but the focus remains on the player with the ball. The style of the video is a fast-paced, action-packed sequence, capturing the excitement and energy of the sport."
+    ],
+    "video_ids": [
+        "DKNyIMlirps_2_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A football player in a blue and green uniform with number 72, holding a football in his right hand, wearing a helmet with a visor.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a blue and green uniform with the number 72, holding a football in his right hand, and wearing a helmet with a visor. All core elements of the description are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a blue and green uniform with the number 72. holding a football in his right hand, and wearing a helmet with a visor. The these elements of the description are present, accurately depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..deb246f470efde7b20860127e42ec739a5b99dd3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d027d7c58286826bd12a5cb1372949bcf4e9a381d65968fd666d5ab073a42b3
+size 268012
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..961cd76860178797d7798bc93e1a682591be65d9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man with a beard is seen enjoying a slice of pizza in a cozy living room setting. He is casually dressed in a blue jacket and a white t-shirt. The man is seated comfortably on a white couch, which is positioned against a white wall. A lamp and a potted plant add to the homely ambiance of the room. The man takes a bite of the pizza, savoring the moment. The video captures the simple pleasure of enjoying a delicious meal in the comfort of one's own home."
+    ],
+    "video_ids": [
+        "-HmlowXOUNU_28_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard, a blue jacket, and a white t-shirt; a slice of pizza; a white couch; a lamp; a potted plant? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard wearing a blue jacket and white t-shirt, eating a slice of pizza while seated on a white couch. In the background, a lamp and a potted plant are visible, matching the described elements. All core objects are present and accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man with a beard, a blue jacket and a t-shirt, sitting a slice of pizza. sitting on a white couch. There the background, there lamp and a potted plant are visible, matching the description elements. The the conditions and present and correctly depicted in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..29d7b587b68123e4262fba6179fd7fed264cdc2a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83b1d3c20aab4f1bfd42d816c94953d4a374f823d3af85ac0816b6a2ceb1077b
+size 154859
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..80b32a5413a209eef371f4eea17344245133bd0b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a bowl of food, which appears to be a type of Asian cuisine, possibly beef and vegetable stir-fry. The bowl is filled with slices of beef, shredded vegetables, and a single orange slice. The food is garnished with a sprig of green onion. The beef is cooked to a medium rare, and the vegetables are bright and fresh. The orange slice adds a pop of color to the dish. The food is served on a white plate, which contrasts nicely with the vibrant colors of the food. The overall style of the video is simple and straightforward, focusing on the food and its presentation."
+    ],
+    "video_ids": [
+        "5MVHFHyZQW0_13_0to172"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bowl of food, slices of beef, shredded vegetables, orange slice, green onion sprig? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a bowl of food topped with slices of beef, shredded vegetables (including bean sprouts and green onions), an orange slice (which appears to be a tomato or similar round red vegetable), and a green onion sprig. All elements described in the condition are present in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a bowl of food that with slices of beef, shredded vegetables,which what sprouts and possibly onions), an orange slice,which appears to be a wedge slice a fruit fruit fruit), and a green onion sprig. The the match in the question are present in the image.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1508ca3644e3f48bef2db247db155bd7df45ee39
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e8ce2514665607b10768b3714b3f09126b40f2ea5ce70537c2d00e5736b7d15
+size 126311
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..c2326223ffdf1e33187886d71147d8943d9a4f4d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a character in a fantasy setting, standing in a snowy landscape with mountains in the background. The character is wearing a dark, armored suit with spikes and a helmet, giving off a menacing and powerful aura. The armor is detailed and intricate, suggesting a high level of craftsmanship. The character is standing with their arms outstretched, as if commanding or controlling something. The overall style of the video is realistic with a touch of fantasy, and the lighting is soft, casting a warm glow on the snowy landscape. The character's pose and the surrounding environment suggest a sense of adventure and exploration."
+    ],
+    "video_ids": [
+        "h6p7da3t1do_1_0to170"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A character in a dark, spiked, and detailed armored suit with a helmet.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a character wearing a dark, spiked, and detailed armored suit with a helmet, which matches the description. The armor is intricately designed with sharp, angular spikes and a helmet that covers the head, fitting the specified criteria.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a character wearing a dark, spiked, and detailed armored suit with a helmet. which matches the description provided The armor is blackately designed with spikes edges angular spikes and a helmet that covers the entire, align the ' '.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..98c8942f19182537bbfddc0a75587abaeb5368b1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be32b8eafe210df2f5581e594d61f6fdc4ae83972d6b90c37040452b085941d4
+size 116207
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..9864b482b61ecba5167e4ecf0768c92dee67f4a3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a basketball game. The main focus is a player from the Memphis Grizzlies, wearing a black jersey with the number 10. He is in the middle of a play, holding the basketball in his hands, ready to make a move. His body language suggests he is focused and determined. The background reveals a bustling basketball court, filled with other players and spectators. The atmosphere is energetic and competitive, typical of a professional basketball game. The video is a snapshot of the intensity and excitement of the sport."
+    ],
+    "video_ids": [
+        "IYfot34F28E_51_81to216"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player from the Memphis Grizzlies wearing a black jersey with the number 10, holding a basketball.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a basketball player wearing a black jersey with \\\"MEMPHIS\\\" and the number \\\"10\\\" clearly visible, holding a basketball. The player's attire and action match the description, even though there is another player partially visible in the frame, which does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a basketball player wearing a black jersey with theMemPHIS\\\" and the number 10\\\" on visible. which a basketball. The player is attire and the match the description provided and though the are a player in visible in the background, which is not contradict the main description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2676a419bc1e3ca1dc1ab86cb53695d0d15a60b0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:92b4a1d1d94dc77fdb93829e4bbc0ce621012aedaaae095f7f5a30d5d2e08950
+size 304529
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..d3ffd07cad3bcdc84965ba930052501f283b58c4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car, specifically focusing on the back seat area. The car appears to be a modern SUV or crossover vehicle, with a spacious and well-designed interior. The back seat is folded down, revealing a flat surface that extends from the back of the front seats to the rear of the vehicle. The interior is well-lit, with natural light coming in through the windows, suggesting that the car is parked outdoors. The car's interior features a combination of black and white colors, with the seats and door panels being white, and the floor and other surfaces being black. The car's design includes a variety of storage compartments and cup holders, and the seats are equipped with headrests and seat belts. The overall style of the video is clean and minimalist, with a focus on the car's interior design and functionality."
+    ],
+    "video_ids": [
+        "NCf-uSDEr8Y_26_84to229"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Back seat (folded down), seats, door panels, floor, storage compartments, cup holders, headrests, seat belts? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows the back seats folded down, revealing the cargo area. It clearly displays the seats, door panels, floor, and storage compartments. Seat belts are visible on both sides, and headrests are present on the rear seats. The video does not contradict any of the listed elements, and all are accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful the back seat being down, revealing the floor area. The also displays the seats, door panels, floor, and storage compartments. The belts are not, the sides of and therests are present on the seats seats. The video does not contradict the of the mentioned objects, and the are visible depicted.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cdb2a9777781c82316fadf81e5e97c75b8e6d486
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1233-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e7e0f3388d0d3213dd547f90404c9191376dd65a1ebdb83f8b659e184407bdd
+size 157199
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..5eabc878e42ad5552a121fee7815990613d4d3fb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a white t-shirt and a baseball cap, standing in a forested area with a clear blue sky. He is smiling and appears to be in a good mood. In the first frame, he is looking directly at the camera, while in the second frame, he is looking to his left, and in the third frame, he is looking to his right. His hand is raised slightly in the third frame, as if he is gesturing or pointing. The forest around him is lush and green, with trees and bushes filling the background. The overall style of the video is casual and friendly, with a focus on the man's expressions and gestures."
+    ],
+    "video_ids": [
+        "GwU1VXsDXbc_48_231to439"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a cap and a t-shirt, speaking and gesturing in an outdoor setting. The subject is consistently identifiable as a man, fulfilling the 'Object(s): A man' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a man wearing a white and a white-shirt, standing and gesturing, a outdoor setting. The presence matches the human as a man, fulfilling the 'Object(s)' A man' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..feb1aa2aa8e93392b5357b22419e0994b18abc39
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fc5b36dde2d9682d73aecab4d8589aaa4488316be273b42f36e4eb5b64e23197
+size 154413
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..f147a02a66b856fdeb0940aa95ae116c96d19c95
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a group of four people, three men and a woman, are standing outdoors, all wearing face masks. The woman is in the center, flanked by two men on either side. The fourth man is standing behind the woman. They are all looking towards the camera, with the woman appearing to be in the middle of a conversation or explanation. The setting is a lush green field with a red car parked in the background. The sky is overcast, suggesting it might be a cool or cloudy day. The overall style of the video is casual and candid, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "N2FmI1FB8yM_31_0to190"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Four people (three men and one woman), a woman in the center, two men on either side, and a fourth man behind her, a red car.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows four people: a woman in the center wearing a pink sari, and three men around her \u2014 one in front of her (partially obscured), one to her right, and one behind her. A red car is visible in the background on the left side of the frame. The composition matches the described arrangement.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows four people, three woman in the center and a black faceash, and three men, her, two on a on her onleftially obscured), one to her right, and one to her. There red car is visible in the background, the left side of the frame. The arrangement and the description scene,}<|im_end|>\n addCriterion",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3295dd936b1229d884e3d94357007d8b71343e87
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e19234eb27af044a82f549e8d34904a2ce95411d71b6e5130db640b5ef0acca
+size 251381
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..36369a9a3ceaf259723e8659b91a37db671838a4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two animated characters, a pink elephant and an orange mushroom, are in a room with a purple wall and a wooden floor. The elephant is pointing at a solar system poster on the wall, which features planets and stars. The mushroom is standing next to the elephant, looking at the poster. The room also contains a pink wardrobe, a yellow toy box, and a green ball on the floor. The characters are standing on the wooden floor, and the overall style of the video is colorful and cartoonish."
+    ],
+    "video_ids": [
+        "GBWi8QJjbyM_267_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two animated characters (a pink elephant and an orange mushroom), a solar system poster, a pink wardrobe, a yellow toy box, and a green ball.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully features the two animated characters (a pink elephant and an orange mushroom), a solar system poster on the wall, a pink wardrobe on the left, a yellow toy box on the right, and a green ball (part of a multicolored ball) visible in the bottom right corner. All core elements from the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video contains ful the two animated characters (a pink elephant and an orange mushroom), a solar system poster, the wall, a pink wardrobe, the left, a yellow toy box on the floor, and a green ball onwhichially the largericolored ball) on on the scene left corner. The elements elements from the description are present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0a91172294d066e41ca16109b1b8a0296c4771a1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:39e47059fec0b61ce03b010d298514e5de55dac722065aa5e1213fd28ead7654
+size 159213
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..03cba38c0f1f1d38bbd01f5be602ca84e599617b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene inside a clear glass bowl filled with water. Two guppies, small, colorful fish, are swimming gracefully within the confines of the bowl. The first guppy, positioned slightly ahead, has a vibrant green body with hints of yellow and orange, while its fins display a mix of brown and black hues. The second guppy, trailing behind, is adorned with a similar color palette but with more prominent red tones on its fins. Both fish exhibit fluid, dart-like movements as they navigate through the water, occasionally creating ripples that disturb the otherwise calm surface. The background is minimalistic, featuring a plain white surface that contrasts with the transparency of the water and the vivid colors of the fish. The lighting is soft and diffused, enhancing the clarity of the water and the details of the fish's scales and fins. There is no significant change or movement in the environment throughout the video, maintaining a consistent and tranquil atmosphere."
+    ],
+    "video_ids": [
+        "b56c2d6d374ea2df0682c4e0dfdc01ea00327e4c90e400be2d7ae8cc77c6a010"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two guppies? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two guppies swimming in a glass container filled with water. Their distinct colorful patterns and fin shapes are visible, matching the description of guppies. The reflection in the water adds visual complexity but does not contradict the core object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two guppies swimming in a bowl bowl. with water. The vibrant colors patterns and shapes shapes are consistent, which the description of guppies. There presence of the water and a interest but does not contradict the core description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cdfcf978f227bdd065c2baa60ee6c5b63d2d9e3c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6434bbbb0c44bfef2c67df53c440ec0cc777a154d4539b25c3a4969d0a0b050
+size 182538
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..48aa0f63ab1d58423cddccef9a596f4be99456d0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a row of canned beverages, likely beer, being dispensed from a machine. The cans are shiny and metallic, with a blue and silver color scheme. The cans are lined up in a row, and the machine is in the process of dispensing them. The cans are being dispensed one by one, and the machine is in the process of opening the cans. The cans are being dispensed into a metal container, which is likely a part of the machine. The machine is likely located in a bar or a similar establishment. The style of the video is a close-up shot, focusing on the cans and the dispensing process. The video does not show any people or other objects, and the focus is solely on the cans and the machine. The video does not show any text or other graphics. The video is likely intended to showcase the dispensing process of the cans, and the focus is on the cans and the machine."
+    ],
+    "video_ids": [
+        "CRcxRMnUbp8_11_0to140"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Canned beverages (beer), machine, metal container? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a machine processing canned beverages, which appear to be beer cans, and includes metal containers as part of the machinery. All elements align with the described conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a machine interacting blue beverages, which are to be beer cans, as a a containers as part of the machinery. The elements align with the ' conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..471576bd680983dac88d7d7032a7f92ad20a8c6f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:feb827ba919f4afb660300e3a42981527f327d5afb8b7e3f2e8cd1c8e8cddf5e
+size 258429
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..b0fa6f9673f6d12d655394ca2007edd29495ef43
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are seen in a mountainous landscape. The man, dressed in a pink shirt, is kneeling on one knee, holding a bouquet of red roses. The woman, wearing a black dress, stands beside him, looking down at him with a smile. The scene is set against a backdrop of a rocky cliff and a wooden fence. The man's proposal is the central focus of the video, capturing a moment of love and commitment in a beautiful outdoor setting."
+    ],
+    "video_ids": [
+        "ACdZia1sNZo_81_48to260"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman as the central subjects. The man is kneeling, and the woman is standing, which aligns with the 'Object(s)' condition of having a man and a woman present. Additional elements such as the landscape and roses do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a man and a woman. the main figures. The man is kneeling and holding the woman is standing, which aligns with the descriptionObject(s)' condition of a a man and a woman in. The elements like as the flowers and the do not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..747be3937d30d57ad2abf81282e7f0d2571c2a54
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:82f8ea0fbcd71955eaba3d0e468a1370c4a4ff6557b7dae23271a2904f72c8ec
+size 207113
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..12c1184f69a82adf9df83a600c8243fce4cd4188
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men in a motorcycle shop. The man on the left is gesturing with his hands, possibly explaining something to the man on the right. The man on the right is wearing a black, white, and red jacket and is looking down at his jacket, possibly checking it. The shop is filled with motorcycle gear, including helmets and jackets. The lighting is bright, and the shop appears to be well-stocked. The men seem to be engaged in a conversation about the motorcycle gear."
+    ],
+    "video_ids": [
+        "GpvjN54NjtU_24_0to182"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, one wearing a black, white, and red jacket who is looking down at his jacket, and another man gesturing with his hands.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men in a motorcycle gear store. The man on the right is wearing a black, white, and red jacket and is looking down at his jacket while buttoning it. The man on the left is gesturing with his hands while speaking. These elements match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men in a store shop store. The man on the right is wearing a black, white, and red jacket and is looking down at his jacket, theing it. The man on the left is gesturing with his hands, talking, The actions match the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a1dde3da1ca260b37cdffaa4c10264b590e6a4b9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d455506deaf10b404d290b7b625735366adbe3b547ed428dbdada8874afba677
+size 198876
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..8d491d53e024790e735de1de1f6eb7a2460104eb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman with blonde hair is seen in a room with a blue door. She is wearing a yellow sweater and has her mouth open as if she is speaking or reacting to something. The room has a clock on the wall and a coat hanging on the door. The woman appears to be the main subject of the video, and her actions suggest a conversation or interaction with someone off-camera. The style of the video is realistic, capturing a candid moment in the woman's life."
+    ],
+    "video_ids": [
+        "6a1gpbDjYAc_0_0to156"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with blonde hair, wearing a yellow sweater. A clock, a coat.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a woman with blonde hair wearing a yellow sweater, which matches the core description. While a coat is visible in the background (likely hanging), it does not contradict the description. The presence of a clock is not clearly visible or identifiable in the video frames provided, but since the core description is largely fulfilled, the answer is 'Yes'.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair wearing a yellow sweater. which matches the description description. Additionally there clock is visible in the background,h hanging on it does not contradict the description as The presence of a clock on also explicitly visible in mentioned in the video, provided, but it the description elements does not met, the answer is 'Yes'.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..caab8baa0bdc740df172e845f050242f1dd3c365
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1234-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1f1a2b847f4b2d2e5a8c3485064b2e5fec7d4f401a6bcb4cc50a8aa7c0f59fce
+size 177735
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..5ab06ee4d19bac90c078d7e139a9b2cec3f5ba75
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is a player wearing a black and teal uniform with the number 27, who is in possession of the football. He is running towards the right side of the frame, with his body leaning forward in a sprint. His helmet is black with a teal visor, and he is wearing white gloves.  In the background, there are other players, some of whom are wearing white uniforms with red accents. They are in various positions on the field, some closer to the camera and others further away. The field itself is a vibrant green, and the sky is clear, suggesting it's a sunny day.  The style of the video is a standard sports action shot, capturing the intensity and speed of the game. The camera angle is slightly elevated, providing a clear view of the player's movement and the surrounding action. The focus is sharp on the player with the ball, while the background is slightly blurred, emphasizing the speed and movement of the scene."
+    ],
+    "video_ids": [
+        "gwDydZmCcyw_43_0to131"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a black and teal uniform with number 27, other players in white uniforms with red accents, and the football.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a black and teal uniform with the number 27, holding a football. Another player in a white uniform with red accents is visible nearby, consistent with the description. The core elements described are accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a player in a black and teal uniform with the number 27, which a football. There player in a white uniform with red accents is visible in, and with the description. The presence elements of are present represented in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..92a03af4b3063df64d6673f737d470593e8f94cc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7d5da7b2bc0f92a457aea43f7eecd42ee03201009d50f3acd4f38958f39ffbc2
+size 249521
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..301cda458208c2e589dc674dde5287bf40ec9627
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a dynamic and stylish advertisement for a BMW car. The car, a sleek and modern model, is captured in motion on a winding road. The car's design is highlighted by its large grille and distinctive headlights. The car is painted in a dark color, which contrasts with the bright and sunny day. The road is surrounded by lush green grass and power lines, adding to the sense of speed and movement. The car's shadow is visible on the road, indicating its direction of travel. The video is shot in a way that emphasizes the car's performance and design, making it an effective advertisement for the BMW brand."
+    ],
+    "video_ids": [
+        "ztGoBzO3x2o_3_0to133"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A sleek and modern BMW car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a sleek and modern BMW car, specifically a silver BMW M5, which matches the description. The car's design, including its aerodynamic lines, distinctive grille, and sporty wheels, aligns with the 'sleek and modern' criteria. The surrounding environment does not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a sleek and modern BMW car, which a model BMW model5, driving is the description of The car is design, including its grilleodynamic shape, large grille, and headlightsy profile, aligns with the 'sleek and modern' attributes. The setting environment, not contradict the description,}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..50352f3b3898afbabe0cb86219d77af01c174cd9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0bdc6af3ac7fc58b2fa1fbf0fb910aebca8c82067f30571cb764b9e8b7ddd9c6
+size 249325
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..dcfa85b832ebc2275ba51d0b220a399bc4175241
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a white horse standing in a grassy field. The horse is wearing a pink saddle blanket and a brown saddle. The horse is also wearing a blue halter. The horse is standing next to a wooden fence. The horse is looking to the left. The grass is green and the sky is blue. The horse is the main focus of the video."
+    ],
+    "video_ids": [
+        "YsYVXRfWVeE_21_185to311"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white horse, a pink saddle blanket, a brown saddle, a blue halter? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a white horse with a pink saddle blanket, a brown saddle, and a blue halter. All specified objects are present and correctly described in the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a white horse wearing a pink saddle blanket. which brown saddle, and a blue halter. The the elements are present and match described in the video content.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ba720d5bdff864658f4111c3b33abaad013c1b0d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:379427f61ad791a62ea050b7cfa539b6306bdcf411a48e27905d50bc83387e4e
+size 162861
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..85055be82da448b1b9f3c7cb15cbab23dd1e30ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two women sitting on a bench, engaged in conversation. The woman on the left is holding a cup of coffee and a cell phone, while the woman on the right is also holding a cell phone. Both women are wearing colorful dresses and headscarves, and they have handbags with them. The bench they are sitting on is blue, and there is a backdrop of a green curtain behind them. The video appears to be a still image, as there is no indication of movement or action. The style of the video is casual and relaxed, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "5OgfsG0SbKc_4_0to116"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two women, one cup of coffee, one cell phone, one cell phone, one dress, one headscarf, one handbag? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two women sitting on a bench. One woman is holding a cup with a straw, and the other is holding a cell phone. Both women are wearing dresses and headscarves. A handbag is visible next to one of the women. All elements mentioned in the description are present, and there are no contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two women sitting on a bench. One woman is holding a cup, a straw, which the other is holding a smartphone phone. Both women are wearing dresses and headscarves. The handbag is visible on to the of the women. The the mentioned in the ' are present, and there are no contradictions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e098a6439621e00d2234a7d194304b3ade1fceae
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f92bc88f2fbdb2fa89eb1cbad4c616e86ac4248eb75291940749119536db934f
+size 139630
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..6cf0dbcb0151ea7fd9ff14ffa72364ddf96b9c16
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a thrilling moment in a football game. The main focus is on two players, one from the Pittsburgh Steelers and the other from the Cleveland Browns. The player from the Steelers, wearing a black and yellow jersey with the number 44, is in possession of the football and is running towards the right side of the frame. The player from the Browns, wearing a white and orange jersey with the number 28, is in pursuit, trying to tackle the player from the Steelers. The action takes place on a football field, with the players' movements creating a dynamic and exciting scene. The video is shot from a side angle, providing a clear view of the players' actions and the intensity of the game."
+    ],
+    "video_ids": [
+        "bS6QNeqL9Po_18_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two players, one from the Pittsburgh Steelers and one from the Cleveland Browns.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two football players in action. The player on the left is wearing a black and yellow uniform with the number 44, which is consistent with the Pittsburgh Steelers' colors and jersey style. The player on the right is wearing a white and orange uniform with the number 28 and the name 'BODDEN' on the back, which matches the Cleveland Browns' team colors and jersey style. Both players are clearly identifiable as representing their respective teams, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts two players players on action on One player on the left is wearing a black jersey yellow uniform, the number 84, which is consistent with the Pittsburgh Steelers' colors. uniform style. The player on the right is wearing a white and orange uniform with the number 83, a Cleveland 'BrowDEN' on the back, which is the Cleveland Browns' colors colors and style style. The players are actively engaged as representing the respective teams, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..51ebb7e8a99e6222fcf3dccfe66b05b15b7e237c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f0b9b4f5e16c182ef89408bc4df8b1278e566d1fa077725e9fb3dc2df5c038c8
+size 274459
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..aba2ca154b6ad8d973740764b88bc3cc93f2c406
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a bartender is seen crafting a cocktail in a bar setting. The bartender is holding a cocktail glass filled with a yellowish drink, possibly a margarita, and is garnishing it with a sprig of mint. The bartender is surrounded by various bottles of alcohol, including tequila, which is likely the base of the cocktail. The bartender is using a cocktail shaker to mix the drink, and a spoon to stir it. The bar counter is well-stocked with a variety of bottles, indicating a wide selection of drinks available. The bartender's focus and attention to detail suggest a professional and skilled approach to crafting cocktails. The overall style of the video is a close-up shot of the bartender and the cocktail-making process, emphasizing the bartender's actions and the drink being made."
+    ],
+    "video_ids": [
+        "hO5-H9X-cH8_0_0to162"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bartender, cocktail glass, sprig of mint, cocktail shaker, spoon, bottles of alcohol (including tequila)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a bartender (a woman in a black top), a cocktail glass with a drink and a sprig of mint, and multiple bottles of alcohol including tequila (labeled 'Tipsy Bartender Tequila'). Although a cocktail shaker and spoon are not visible in the frames, the core elements requested are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as showing showing a bartender performingd person in a black apr and a cocktail glass with a spr and a sprig of mint, a a bottles of alcohol, whatquila.thoughabeled asTey Bartender').quila'). The the cocktail shaker is a are not visible in the frame provided the bartender elements of are present, the depicted.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..12b0e3c783f217dc2cfc8d9f21aa20641834f76c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3212ed47ca1414fcf4e19a60c3e40fe19da4079a7ff6db20db9f411bbfd455f
+size 168638
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..93fd81d34543f18deeb0e5e89ed190ef7a16c201
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a black Mercedes-Benz sports car driving down a street. The car is sleek and shiny, with a large grille and the Mercedes-Benz logo prominently displayed. The car's headlights are on, and it appears to be moving at a moderate speed. The street is lined with buildings, and the car is driving on the right side of the road. The sky is clear and blue, suggesting it's a sunny day. The car's occupants are not visible in the video. The style of the video is realistic, with no visible special effects or animation. The focus is on the car and its movement, with the surrounding environment serving as a backdrop."
+    ],
+    "video_ids": [
+        "B5-5hP2NOdE_11_0to112"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black Mercedes-Benz sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a black Mercedes-Benz sports car, specifically a SLS AMG model, which matches the description. The car is shown in motion and stationary, with clear visibility of its design, logo, and color. There are no elements that contradict the core description of the object.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a black Mercedes-Benz sports car, which a model-Class AMG model, as is the description of The car is shown in motion, the positions with clear visibility of its front elements including, and grille. The are no elements in contradict the description description of the object.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cbd81d4e196292503605fc748cf1ad7b2c81f91e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af2546d20d9818ad17d3b47629059b14e74a2dfbb420859014ab4b9fe4b3f41a
+size 236607
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..074aa54e95791fd3bcadf5cd71743576a435a4f0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young man is seen shopping in a store. He is wearing a red hoodie and is pushing a shopping cart. The store is well-lit and has aisles filled with various items. The man is looking at the camera, possibly taking a selfie. In the background, another person can be seen, also shopping. The overall style of the video is casual and candid, capturing a typical day of shopping."
+    ],
+    "video_ids": [
+        "B1r9aX25uog_2_84to215"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man, a shopping cart, another person shopping? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man in a red hoodie sitting in a shopping cart, which matches the 'Object(s)' condition. Another person is visible in the background, also shopping, which aligns with the description. The setting is clearly a retail store with aisles and shopping carts, supporting the context of shopping. No elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man in a red hoodie, in a shopping cart, which ful the 'shopping(s)' condition. Additionally person is also in the background, also shopping, which aligns with the '. The setting appears a a store store, shelvesles and shelves carts, further the context of the. The elements contradict the given description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3817fa05f7949f6a2d6909bb1ed2da813995a40c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1235-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de75c5fe3986af653d5fc2925e6680616b2746173a1eb8042a5bc8db3c17d3f7
+size 222040
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..f86f41cd012f9eb42aa71e2a082f4fe0e7a3df00
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young woman with a headscarf, sitting in a chair and engaging in a conversation. She is wearing a gray sweatshirt with a white stripe and has a lanyard around her neck. The setting appears to be an indoor space with a blurred background, possibly a room with a desk and a chair. The woman is gesturing with her hands as she speaks, indicating an active discussion or explanation. The style of the video is casual and candid, capturing a moment of interaction between the woman and another person, who is not visible in the frame."
+    ],
+    "video_ids": [
+        "QFxpu1erWo0_33_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Young woman with a headscarf, wearing a gray sweatshirt with a white stripe, and a lanyard around her neck.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman wearing a headscarf, a gray sweatshirt with a white stripe, and a lanyard around her neck, which matches the description. The presence of additional elements like the chair and background does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman wearing a headscarf, a gray sweatshirt with a white stripe, and a lanyard around her neck. which matches the description provided The presence of a elements like the car and the does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..629e61c85707d1d4a91a973e1d7fb86b40c0ad9c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ebdcb5384278f6632a10d8ac6968b9321d55e7a06aa92c2dce9288324e4ccbc7
+size 202466
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..6dab53699375484218c3f0b95dc1db9cbc712388
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car from the perspective of the driver's seat. The car has a modern design with a sleek dashboard and a touch screen display in the center console. The steering wheel is on the right side of the car, indicating that it is a right-hand drive vehicle. The seats are upholstered in a dark gray fabric, and there is a red stitching detail on the seat edges. The car's interior is well-lit, with ambient lighting coming from the dashboard and door panels. The car appears to be a luxury vehicle, given the high-quality materials and attention to detail in the interior design. The video is likely a promotional video for the car, showcasing its interior features and design."
+    ],
+    "video_ids": [
+        "EEXSj4s3gJY_3_17to175"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Modern car, sleek dashboard, touch screen display, steering wheel, seats upholstered in dark gray fabric, red stitching detail on seat edges? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a modern car interior with a sleek dashboard, a touch screen display, a steering wheel, and seats upholstered in dark gray fabric. Red stitching detail on the seat edges is visible, matching the description. All core elements are accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts the modern car interior with a sleek dashboard, a touch screen display, a steering wheel, and seats upholstered in dark gray fabric with The stitching detail is the seat edges is also, matching the description. The elements elements of present represented in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1383180407c49f126fb4b6516b71b0a450dc1fcb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:05be39c8ca95576dd66e8034a668f577ba7ad91c13d6300106aa49475a32196c
+size 203691
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..1e7b46e4f2572222f332b8d5c350e694027c39ec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene sunset scene at a coastal location. The sun is positioned low on the horizon, casting a warm golden glow across the sky and reflecting off the calm ocean surface. The sky is partially cloudy, with streaks of light breaking through the clouds, adding depth to the scene. The coastline features a rocky cliff on the right side, where a building is perched, silhouetted against the vibrant backdrop. The beach in the foreground is sandy and appears relatively empty, with gentle waves lapping at the shore. The water is a deep blue, contrasting beautifully with the golden hues of the sunset. Throughout the video, there are subtle movements of the waves and the occasional bird flying across the sky, enhancing the tranquil atmosphere. The overall ambiance is peaceful and picturesque, highlighting the natural beauty of the coastal landscape during the golden hour."
+    ],
+    "video_ids": [
+        "6678f23838fbcc8f0fcaa2e9898f34b9a595c11cccfe5897fd241f3fbe28e23e"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Sun, clouds, rocky cliff, building, beach, waves, birds? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts all the specified elements: the sun is visible setting over the horizon, casting a golden glow on the water; clouds are present in the sky; a rocky cliff with a building on top is visible on the right; a beach with waves gently rolling in is in the foreground; and birds can be seen flying in the sky. All core elements are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful the the mentioned objects: the Sun setting visible setting over the horizon, there a warm glow over the water and there are present in the sky, there rocky cliff is a building on top is seen; the right side the beach with a crashing l in is shown the foreground; and there can be seen flying in the sky, The these elements are present and the represented.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..62f98093ec707bfb8dbf335f93692b9380de52d0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f5ee8ee6832521d6df9367dc42f9cd0bb383d44e2313a7fa12f8bf57507e328
+size 83963
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..7543ba31c5c0b5d9d49d8808ccccdbaaef04c688
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a large, brown and black moth with a distinctive white and black eye spot on its wing. The moth is perched on a branch, surrounded by green foliage. The moth's wings are spread wide open, revealing its intricate wing patterns. The moth appears to be in a state of rest, with its antennae relaxed and its body still. The video is shot in a realistic style, with a focus on the moth and its immediate surroundings. The camera angle is slightly elevated, providing a clear view of the moth and the branch it is perched on. The lighting in the video is natural, with the moth and the foliage bathed in soft, diffused light. The overall mood of the video is calm and serene, with the moth appearing peaceful and at ease in its natural habitat."
+    ],
+    "video_ids": [
+        "hQZCI3NOygU_96_122to267"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large, brown and black moth with a white and black eye spot on its wing.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large, brown and black moth with a distinct white and black eye spot on its wing, which matches the description. The moth is the central focus, and although there are other elements like a praying mantis and background foliage, they do not contradict the core description of the moth.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close moth brown and black moth with a white white and black eye spot on its wing. which matches the description provided The moth is per central focus of and the there are additional elements like a plant mantis and a foliage, they do not contradict the core description of the moth.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5505059f7fe8e566874ab01f39094ae45837a65a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2583f76917aad40cd79701bf0082eb979c9f9647615c6def4157607144ed6336
+size 154672
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..4ec2381e2bb7141c3edcee3bdc92896ee1090e44
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a close-up view of a culinary dish placed on a wooden surface. The dish appears to be a type of sushi or sashimi, featuring a white, translucent piece of fish or seafood at its center, garnished with a small green leaf. Surrounding this central piece are two brown, glossy, and slightly translucent pieces of food, possibly another type of fish or a different ingredient, arranged symmetrically on either side. The wooden surface beneath the dish has visible scratches and marks, indicating frequent use, likely for food preparation.  As the video progresses, there is a noticeable change in the appearance of the dish. The white piece of fish or seafood begins to melt or dissolve, spreading outwards and blending with the surrounding brown pieces. This melting process is gradual and continuous, creating a smooth transition between the different components of the dish. The green garnish remains intact throughout the transformation, adding a pop of color against the melting backdrop. The overall effect is one of a delicate and artistic"
+    ],
+    "video_ids": [
+        "3beaf787aa1e15bd73f17dec7b593e52119f5c35ed23d66fd54d7a71516f4fc2"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white, translucent piece of fish or seafood (central), two brown, glossy, and slightly translucent pieces of food (surrounding the central piece), a small green leaf (garnish).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a central white, translucent piece that resembles fish or seafood, flanked by two brown, glossy, and slightly translucent pieces, with a small green leaf garnish on top. These elements are clearly visible and match the described composition, even though the video includes a brief overlay of a 'thumbs up' icon, which does not contradict the object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a central white, translucent piece of resembles a or seafood, surroundedanked by two brown, glossy, and slightly translucent pieces of which a small green leaf asish on top. The elements match consistent present and match the description conditions.\"\n fulfilling though the video is a wooden moment of a woodenloading up' icon, which does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..98e611d11b4d87dff0a51e54902635ac80a287ef
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e23628c6009bcbdb26ee772e3fbcfbe9c92abda81865709073a12c928270c846
+size 54653
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..21a4fd89db0472410d3508cc5991ef95cd57daf2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with blonde hair, wearing a black top and large earrings. She is seated in a garage-like setting with a truck in the background. The woman appears to be speaking, and her facial expression suggests she is engaged in a conversation or interview. The lighting in the garage is bright, and the overall style of the video is casual and candid."
+    ],
+    "video_ids": [
+        "3gVdM8k_V2U_29_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with blonde hair, wearing a black top and large earrings.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair, wearing a black top and large earrings, which matches the description. The background elements, such as the truck and tires, do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a woman with blonde hair, wearing a black top, large earrings. which align the description provided The background appears, such as the car and the, do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dbf444f86629edd9adfb6e51368230158452c50c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9b8bda1fd34b0acf8ff8b9844127cfc9f7db568210cddd05817bb50d8f928233
+size 212680
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..3eb9f7cfd5676e8f051c2599639e221c6b9f9776
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a golf club, specifically a driver, being held in a person's hand. The club is black with a blue and white logo on the head. The person is standing on a green carpet, and there is a blue mat in the background. The style of the video is a simple, straightforward product demonstration, focusing on the golf club and its design. The lighting is bright, highlighting the details of the club's head and grip. The background is minimalistic, ensuring that the viewer's attention remains on the golf club."
+    ],
+    "video_ids": [
+        "837An20fbt8_38_0to113"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black golf driver with a blue and white logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black golf driver with a blue and white logo ('ST 1906' with blue accents and white lettering). The object is clearly visible and matches the description, with no conflicting elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black golf driver with a blue and white logo,7'),71'),'),'), or a and). a texting). The object matches held a and matches the description provided with no additional elements present would the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c9dae50051c35cd411e07ea5557343d36b15cb8f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:754544f580ac982c5aea774d8c11ef82bee7ac7f6a649fbd5e30a595295d7351
+size 118725
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..d83eef2e937b8aac972d70b67bf513a8d14abfc3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen standing in a park-like setting with trees and a playground in the background. She is wearing a colorful, sleeveless dress with a floral pattern and a necklace. Her hair is styled in an updo, and she has a serious expression on her face. The lighting suggests it is daytime. The style of the video is realistic, with a focus on the woman and her surroundings."
+    ],
+    "video_ids": [
+        "Dp1DFvQTcJk_115_0to144"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a colorful, sleeveless dress with a floral pattern, a necklace, an updo hairstyle, a serious expression.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a colorful, sleeveless dress with a floral pattern, a necklace, and an updo hairstyle. Her expression appears serious. All core elements of the description are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a colorful, sleeveless dress with a floral pattern, a necklace, and an updo hairstyle. Her expression appears serious. The the elements of the description are present, accurately depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b18c112659d067c0aa2310c42d84996584a99e7e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1236-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:70acbf52bd607356a6695dcdb89fc3c664ae1aea373de716d28f64178fe52b00
+size 185794
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..2532c89bf5390619228beddd529c25b4ceaac983
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a vintage car being worked on in a garage. The car is an orange and black vehicle with a black hood that is currently open, revealing the engine. The car is parked on a lift, which is a common piece of equipment in a garage used to raise vehicles for maintenance and repair work. The garage itself is spacious and well-lit, with various tools and equipment visible in the background. The style of the video is straightforward and documentary-like, focusing on the car and the work being done on it without any additional narrative or embellishment. The video likely serves as a visual record of the car's maintenance or restoration process."
+    ],
+    "video_ids": [
+        "3_sHJG7jfEo_17_344to466"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A vintage orange and black car with a black open hood revealing the engine, a lift? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a vintage orange and black car with its hood open, revealing the engine, and the car is positioned on a lift. The scene is set in a garage, and while other vehicles and equipment are visible in the background, they do not contradict the core description. The main subject\u2014the car on the lift with its hood open\u2014is clearly depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a vintage orange and black car with a black open, revealing the engine, and it car is on on a lift. The description is set in a garage, which there there cars are equipment are visible in the background, they do not contradict the main description of The presence focus, vintage with the lift\u2014 the hood open\u2014is clearly depicted,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..48474c38c60b8ab87e930b5365c6d7891b3192b7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c0e7398ecf0a995b4429d89bed1b00dd6f8a31b2ce866ee457a3a89e5a9c26e
+size 279830
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..dd5d9aa8049a8c7a1e0d5af43e08b8a06145f853
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a red car's engine compartment. The engine is visible with various components such as the air filter, spark plugs, and fuel lines. The car's hood is open, revealing the engine's intricate details. The style of the video is a time-lapse or a series of still images, capturing the different parts of the engine in a sequential manner. The focus is on the engine's components and their arrangement, providing a detailed view of the car's internal workings. The video does not include any text or additional elements, keeping the viewer's attention solely on the engine."
+    ],
+    "video_ids": [
+        "wbTl8RXvONE_155_16to144"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Engine, air filter, spark plugs, fuel lines? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a detailed view of a red car's engine, including components such as the air filter (the silver rectangular cover with circular openings), spark plugs (the cylindrical components on the engine cover), and fuel lines (the yellow hose and other metallic tubes connected to the engine). All specified elements are present and visible, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows an car view of a car car's engine bay which components such as the air filter,visible cylindrical circular part), a patterns), spark plugs (the small objects with the engine),), and fuel lines (the black and visible other black tubes). to the engine). The these objects are present and identifiable in fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..521630f1eb2af0b781eabe6e2d6770fb9f97f60e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26c15fe9340a054292c8905756f8c3b0e4e431e472e2e4288df9843d5fc48960
+size 173231
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..944bd24e81e7fe4787e55a9d2f128ebb7c438558
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a whimsical scene where a small orange van is being loaded onto a wooden dock by two animated characters. The van is parked on the dock, and the characters are standing on either side of it, pushing it towards the water. The dock is situated in a body of water, and there is a small boat with a green flag on it nearby. The characters are dressed in white hats and black jackets, and they appear to be working together to load the van onto the boat. The scene is rendered in a cartoon-like style, with bright colors and exaggerated features. The overall atmosphere of the video is playful and lighthearted."
+    ],
+    "video_ids": [
+        "TO9czJfNozQ_25_0to167"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small orange van, two animated characters, and a small boat with a green flag.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small orange van on a makeshift raft, which is being pushed off a dock by two animated characters dressed as sailors. The raft has a green flag (made of green material) attached to a pole, fulfilling the core description. While the raft is not a traditional boat, it functions similarly in this context, and no elements contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a small orange van, a dock dock, two can a used by a dock. two animated characters. in chefs. There van is a small flag attachedlikely of a paper) attached to it pole. and the ' description. The there setting is not a traditional boat, it serves as in the context, and the additional contradict the given.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a66b3a7bdbc2b5fd59a2ba53e568e042dc0010b0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f9d79e7e4239bfc34d7eb76e7b324d05648c6210492f6c77d88e07e4e3ab7ba
+size 350526
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..11d4b3300a0177fe8728baa5f3c4bf88e032d061
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a compilation of three frames featuring a football player in action. The player is wearing a white and green uniform with the number 12, and a yellow helmet with a green logo. The player is holding a football in his right hand, and his body language suggests he is preparing to throw the ball. The background shows a blurred stadium filled with spectators, indicating that the scene is taking place during a football game. The style of the video is dynamic and action-oriented, capturing the intensity of the sport. The focus is on the player and his actions, with the background serving to provide context and atmosphere."
+    ],
+    "video_ids": [
+        "JgmvU1kG_7I_4_119to244"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A football player in a white and green uniform with the number 12, a yellow helmet with a green logo, and a football in his right hand.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a white and green uniform with the number 12, a yellow helmet with a green logo, and holding a football in his right hand. These elements match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a white and green uniform with the number 12, a yellow helmet with a green logo, and a a football in his right hand. The elements match the description provided, the questionObject(s)' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8b49b54db8cac567e667112298cae8c0f17b4851
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb977258c5be4705256fab88c4509a86cfde0282f2d4f74f2891ca394aec0d8c
+size 351398
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..a52b3734cfbdd0a4eb1d83e42fc59f0fcbad3107
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man with a beard and glasses, wearing a colorful striped shirt and a plaid cap, standing on a wooden boardwalk by a river. He has a blue backpack and a camera around his neck. In the background, there are buildings and a crane, suggesting an urban setting. The man appears to be enjoying his time outdoors, possibly on a trip or a vacation. The style of the video is casual and candid, capturing a moment in the man's day."
+    ],
+    "video_ids": [
+        "8d0ioLYPj6Q_21_0to174"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and glasses, wearing a colorful striped shirt and a plaid cap, a blue backpack, and a camera around his neck.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and glasses, wearing a colorful striped shirt, a plaid cap, a blue backpack, and a camera around his neck. These features match the description exactly. Additional elements like the background scenery and another person walking do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and glasses, wearing a colorful striped shirt and a plaid cap, and blue backpack, and a camera around his neck. The elements match the description provided, The elements such the dock with, the person in in not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..05c22fa77b0e938e7a00de1e6c8e5e1821964888
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f0459da808d9c040ae5d6c2d9ee222fea19ff04dc44e1c5f495b64525f1dc051
+size 243232
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..001962b27a6279b344d87f80ba029d37a63995da
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young girl with blonde hair and pink pigtails is seen standing in front of a wooden door. She is wearing a pink dress with white polka dots and is holding a pink balloon with white stars. The balloon is floating above her head. The door is open, revealing a garden with a wooden fence and a path leading to a gate. The girl appears to be looking out of the door, possibly waiting for someone or something. The overall style of the video is cartoonish and whimsical, with bright colors and a playful atmosphere."
+    ],
+    "video_ids": [
+        "iLJxYqnGY84_129_61to192"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young girl, a wooden door, a garden, a wooden fence, a path, and a gate.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by showing a young girl holding a balloon, standing in front of a wooden door. The background includes a garden with a wooden fence, a path, and a gate. Although the scene briefly transitions to an ambulance in a forest, the core elements requested are present in the initial frames and do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting a young girl holding a balloon, a in front of a wooden door. The background includes a garden with a wooden fence and a path, and a gate. The the girl is shows to show outdoor, the different, the core elements of in present in the initial frames.\"\n the not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f191c5398444dc671a044ead844f38eb497d6cee
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:36fd0050654bc1e51bdcd9db545aeaa28e11c4781f9091d1ca84f180baaa3b16
+size 229614
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..cb1e2e271e0f53b5d98536587dfc813bdcc81039
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a group of Buddhist monks in a serene setting. The monks, dressed in traditional orange and red robes, are seated in a line, their bald heads reflecting the light. The monks are engaged in a peaceful activity, their hands gently folded in their laps. The camera pans across the line of monks, capturing their calm demeanor and the tranquility of their surroundings. The monks' robes contrast beautifully with the natural light, creating a visually striking image. The video is a testament to the monks' dedication to their spiritual practice and the peacefulness of their environment."
+    ],
+    "video_ids": [
+        "2mafx0FKHcc_35_0to113"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Group of Buddhist monks, traditional orange and red robes.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group of Buddhist monks seated together, wearing traditional orange and red robes, which aligns with the description. There are no elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a group of individuals monks wearing in, wearing traditional orange and red robes. which aligns with the description provided The are no additional in contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2e06f36786b76df1ccbb9a12e82cdde9c0d8f3e0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb5f1e8eeb2baf3e248c4bbdbbe4a58110c75e48019d34b37886026b843cc7f0
+size 165890
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..2e0db2d9a0cb975acc0c65dd91050682e2942637
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen in a kitchen, making a heart shape with his hands. He is wearing a blue shirt and a red hat. The kitchen is equipped with a refrigerator, an oven, and a sink. The man is standing in front of the refrigerator, which is located on the right side of the frame. The oven is situated on the left side of the frame, while the sink is located in the middle. The man's action of making a heart shape with his hands suggests that he might be preparing to cook or bake something. The overall atmosphere of the video is casual and homely."
+    ],
+    "video_ids": [
+        "IZMLV4nEWsU_13_408to569"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a refrigerator, an oven, a sink, a blue shirt, and a red hat.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a blue shirt and a red hat (used as a blindfold), standing in a kitchen. A refrigerator and an oven are visible in the background. While a sink is not prominently visible, it is implied to be present in the typical kitchen setup. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing a blue shirt and a red hat,which as a visfold). standing in a kitchen. The refrigerator, an oven are visible in the background, The the sink is not directly featured in it is implied to be part in the kitchen kitchen setting. The the elements of in present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a8fc627a753fccc7d94df1b4b06e2cbd6f698dce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1237-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:499e95f0dbf0ee08a51728a9acb127bcfaf0b964fa7821427c6435524e0cad5b
+size 102950
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..7ff1dd25409095a35a6830374a751079b38547ba
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car, specifically focusing on the driver's side. The car appears to be a modern SUV, with a sleek design and a focus on comfort and functionality. The steering wheel is prominently displayed, featuring the Honda logo in the center. The dashboard is equipped with a touch screen display, which is likely used for navigation and entertainment purposes. The car's interior is well-lit, with natural light coming in through the windows, suggesting that the car is parked outdoors. The seats are upholstered in a light-colored material, and there are cup holders visible in the center console. The overall style of the video is straightforward and informative, providing a clear view of the car's interior features."
+    ],
+    "video_ids": [
+        "jo9WSXyLdWc_47_0to146"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, Honda logo, touch screen display, seats, cup holders? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the steering wheel with the Honda logo prominently displayed in the center. A touch screen display is visible in the center console, and both front seats are visible. Cup holders are also present in the center console area. All specified objects are accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a interior wheel, the Honda logo, displayed in the center. The touch screen display is visible on the dashboard console, and the the seats are visible, The holders are also present, the center console.. The the objects are present depicted in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..db1daa10db8450581722b900b3c928924ccaef7e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e62f470bb5807f179ce22308437e20205c55af738d1c7388a474cb6e23ed7d23
+size 178033
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..0af4326a90fea513f618b1108ec130e601788e4c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a parking lot with three sports cars parked in a row. The first car is a red Jaguar, the second is a blue Porsche, and the third is a red sports car. The cars are parked in a parking space with white lines. In the background, there are trucks and a building. The style of the video is a simple, straightforward documentation of the scene without any additional effects or music. The focus is on the cars and their colors, with the background serving as context for the location."
+    ],
+    "video_ids": [
+        "bWQ_2HG8IWg_50_0to191"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three sports cars (a red Jaguar, a blue Porsche, and a red sports car).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three sports cars parked in a row: a red Jaguar (left), a blue Porsche (middle), and a red Mercedes-AMG GT (right). Although the description mentions 'a red sports car' for the rightmost vehicle, it is actually a red Mercedes-AMG GT, which is still a sports car. The core description is largely accurate, and additional elements like other cars and trucks in the background do not contradict the primary focus.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows three sports cars, in a row, a red car,or), a blue Porsche (middle), and a red sports-BenzAMG GT (right). The the Mercedes mentions aa red sports car' instead the thirdmost vehicle, it is actually a Mercedes Mercedes-AMG GT, which is a a sports car. The presence elements of largely matched, and the elements ( the vehicles and a in the background do not conflict the main focus on}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f7ab26a512e6046c5cd2354c99a01a50e47e285d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77880828ea169888071f972acf27124e7f385353cc30f05f013e4adcbaeb907b
+size 98742
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..0b50a7d900ab7170459ea2bd401cf04dc100e4ef
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in a cluttered room, wearing a black t-shirt with a humorous message. He appears to be speaking or presenting, as suggested by his closed eyes and the microphone attached to his shirt. The room is filled with various objects, including boxes, shelves, and a statue, indicating a workspace or storage area. The style of the video is casual and informal, with a focus on the man and his surroundings. The lighting is natural, and the camera angle is slightly elevated, capturing the man from the chest up. The overall impression is of a relaxed, informal setting, possibly a workshop or a personal space."
+    ],
+    "video_ids": [
+        "4lp3J0L9TNI_79_39to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black t-shirt with a humorous message, a microphone attached to his shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The man in the video is wearing a black t-shirt with the humorous message 'THIS PERSON ACCEPTS HIGH FIVES FROM STRANGERS', and there is a small microphone clipped to his shirt. These elements match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video in the video is wearing a black t-shirt with a text message 'THE TO ISS NO RIVE' STRANGERS' which there is a microphone microphone attached to his shirt. The elements match the description provided, the questionObject(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9d81dc2637b2d56ad26c17f9df390fe4f08c1a2c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f1011f9545e70e7840a370a5b9ef2862454f3210279c8699f37273fd80a8efc0
+size 182583
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..ddc7257cebefdabc44e1744110b88c3a9a88cec5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man enjoying a meal at a restaurant. He is wearing a black t-shirt with a white text that reads \"If it's not spicy, I'm not eating!\". The man is using chopsticks to eat a dish that appears to be a type of sushi. The restaurant has a casual and cozy atmosphere, with wooden tables and chairs. The man is seated at a table with a bowl of soup in front of him. The background shows other patrons enjoying their meals, and the restaurant has a warm and inviting ambiance. The video is shot in a realistic style, capturing the everyday life of a person enjoying a meal at a restaurant."
+    ],
+    "video_ids": [
+        "KzZmPXGlMRI_72_0to156"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a bowl of soup, a dish of sushi, and other patrons.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man seated at a table in a restaurant, holding chopsticks and eating. A bowl of soup and a dish of sushi are visible on the table in front of him. Other patrons are also visible in the background, seated at nearby tables. All core elements mentioned in the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting at a table with a restaurant, eating chopsticks and eating from In bowl of soup is a dish of sushi are visible on the table in front of him. The patrons are also present in the background, sitting at tables tables. The elements elements of in the description are present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c5f2c633985f3c7b9558c05e9f7acbf8cd2b37af
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a591ea72ff5d0fa7813ad81d5a172f7e423f74b08b3a9949664e07f29bfa1d71
+size 115118
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..2a7bfcee663d006f1ac9ee9bf4bacd070c9e4fd6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse of a garage filled with various vehicles and bicycles. The garage has a high ceiling with exposed beams and ductwork. The style of the video is a realistic, documentary-style time-lapse, capturing the changes in the garage over time. The garage is filled with a variety of vehicles, including cars, motorcycles, and bicycles. The vehicles are parked in a somewhat haphazard manner, with some vehicles parked in front of others. The bicycles are hung from the ceiling, adding a unique touch to the garage. The garage is well-lit, with natural light coming in through the windows. The video captures the garage in different times of the day, with the lighting changing as the sun moves across the sky. The garage appears to be a space for both storage and display of vehicles and bicycles."
+    ],
+    "video_ids": [
+        "EU-yHC_2Z8Y_22_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Various vehicles (cars, motorcycles) and bicycles.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple vehicles including cars (a green vintage car in the foreground, a black classic car with a white roof, and a red car in the background), and bicycles hanging from the ceiling. There are no motorcycles visible, but the presence of cars and bicycles fulfills the 'various vehicles (cars, motorcycles) and bicycles' condition since motorcycles are not required to be present for the condition to be met.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a vehicles, cars anda blue and car and the foreground and a black car car, a red roof in and a sleek car in the background) and bicycles hanging from the ceiling. This are no motorcycles visible in but the presence of cars and bicycles fulfills the 'Variousious vehicles'cars, motorcycles) and bicycles' condition. motorcycles are not explicitly to be present for the condition to be met.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..156fd98914303994e5f75881dde7f6b386bbfc50
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be62085d355250c81a991526ecfe84def6d397606180493a4bf47677ec85c4dc
+size 323082
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..3e0c0cafffcbcdf7b0f88ab1a323ebeac8b0ed61
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a bowl of food, likely a soup or stew, with a focus on the ingredients. The bowl is filled with a dark, rich broth that has a glossy appearance, suggesting it might be a thick or creamy soup. The main ingredient appears to be pieces of chicken or another type of poultry, which are cooked and have a golden-brown color. The chicken is garnished with chopped green onions, which are sprinkled on top, adding a pop of color and likely a fresh flavor to the dish. The green onions are bright green and look fresh, indicating they were likely added just before filming. The style of the video is simple and straightforward, focusing on the food without any additional context or background. The lighting is bright and even, highlighting the textures and colors of the ingredients without creating harsh shadows. The video does not contain any text or additional elements, keeping the viewer's attention solely on the food. The overall impression is that of a home-cooked meal, possibly a family recipe, presented in a way that emphasizes the care and attention that went into its preparation."
+    ],
+    "video_ids": [
+        "GwtdiGinctc_6_0to156"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chicken pieces (cooked and golden-brown), chopped green onions (bright green, fresh).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bowl of cooked chicken pieces that appear golden-brown and are coated in a sauce. The chicken is garnished with bright green, freshly chopped green onions. These elements match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows pieces dish of food chicken pieces that appear golden-brown, are garn in a sauce. There chicken is garnished with chopped green, fresh chopped green onions, The elements match the description provided, the questionObject(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7684abef1d8e6103c7f861bb92d5a123cea3439d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e531405815840c16ac726b1e6c326412ba80b0dfcdfafdbfdcde40ab9098b5b8
+size 116620
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..db037d8e9ce39c3cd467007ffeb06dea76293579
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the growth of a red mushroom in a forest setting. The first frame shows the mushroom in its early stages, with a small cap and a slender stem. In the second frame, the mushroom has grown significantly, with a larger cap and a thicker stem. By the third frame, the mushroom has reached its full size, with a vibrant red cap and a sturdy stem. The mushroom is surrounded by fallen leaves and twigs, indicating that it is growing in a natural, wooded environment. The video provides a close-up view of the mushroom, allowing viewers to observe its growth and development in detail. The style of the video is realistic and naturalistic, capturing the beauty and complexity of the mushroom's growth in its natural habitat."
+    ],
+    "video_ids": [
+        "qdslTdBxBeA_2_0to190"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red mushroom? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features red mushrooms growing among fallen leaves and twigs, which matches the description of 'A red mushroom'. The mushrooms are clearly visible and are the main focus of the video, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a mushrooms with in leaves leaves. grassigs. which align the description of 'A red mushroom'. The presence are the visible and the the main focus of the video, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d46065f79a0212f6014cebeadaf5a16838394b2e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:371bc58a6ef5f37f8ed8d07cb016f4cf96f845501ba4df1429ac9ce3349cb69e
+size 74612
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..f264968635dfa7461734cffb7e9912ab72bd7529
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man sitting in a black chair, drinking from a black cup with a red stripe. He is wearing a black hoodie and has a beard. The man is looking to the left, and his expression is serious. In the background, there is a man wearing a suit and tie, and he is looking down at a laptop. The setting appears to be a professional environment, possibly an office or a conference room. The style of the video is candid and unposed, capturing a moment in the man's day."
+    ],
+    "video_ids": [
+        "DeYwii5A7LY_33_0to155"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in black chair, Man in suit and tie, Black chair, Black cup with red stripe, Black hoodie, Laptop? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a black hoodie sitting in what appears to be a black chair, holding a black cup with a red stripe. In the background, there is a man in a suit and tie, and a laptop is visible behind him. All elements described in the condition are present in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man in a black hoodie sitting in a appears to be a black chair, holding a black cup with a red stripe. Another the background, there is another man in a suit and tie, and a laptop is partially on him. The these in in the condition are present in the video,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..49f6aff4f0aeab786b9a72125941a1aa5043a34f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1238-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5610ec341c4c70972cc8af4242be1eb1b9e51678a041f48d0268e0f3da73893a
+size 119610
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..22d45e9039b1fad49fa3104fa9a096a036b3f97c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a fire pit with a few ears of corn being cooked. The corn is placed on top of the hot coals, and the fire is visible beneath the coals. The corn is being cooked evenly, with the kernels turning a golden brown color. The fire pit is made of metal and has a handle on the side for easy use. The coals are glowing red and are covered with ash. The fire pit is placed on a wooden surface, and the background is blurred, focusing the viewer's attention on the corn and the fire. The style of the video is a simple, straightforward cooking tutorial, with a focus on the process of cooking corn on the cob over an open fire."
+    ],
+    "video_ids": [
+        "2ehJ86pbwvo_17_0to113"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A few ears of corn, hot coals, and a metal fire pit.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a few ears of corn being grilled over hot coals inside a metal fire pit. The corn is visibly charred, and the coals are glowing, indicating they are hot. The metal structure of the fire pit is also visible, confirming the core elements described in the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a few ears of corn placed cooked over hot coals in a metal fire pit. The presence is positioned yellowred, indicating the fireals are glowing red indicating they are hot. The presence fire of the fire pit is also visible, fulfilling the presence elements of in the question.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b8bf3b90a779f763796fd9c324a338a7a043db8e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:53f18f0aff46f170ec4633b0ded244341a709655e6230fc33bb18ad79b0a6091
+size 183451
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..8f1111b071c25f708a55bec5fe420279ea2fccfa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a close-up shot of a clear glass filled with blue liquid and ice cubes. A yellow rubber duck is floating on top of the liquid, positioned centrally within the frame. The background is entirely black, which creates a stark contrast with the brightly colored rubber duck and the blue liquid. The lighting is focused on the glass, highlighting the texture of the ice cubes and the smooth surface of the liquid. The rubber duck remains stationary throughout the sequence, with no noticeable movement or change in its position. The overall scene is simple yet visually striking due to the vibrant colors against the dark backdrop. There are no other objects or characters present in the video, and the camera maintains a steady, fixed position, ensuring that the focus remains solely on the glass and its contents."
+    ],
+    "video_ids": [
+        "63bb6eda9b18960159fefcbed300c14589615df05247075428af86af8af47621"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A yellow rubber duck? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a yellow rubber duck floating in a glass of blue liquid with ice cubes, which matches the core description of the object. Additional elements like the glass, ice, and blue liquid do not contradict the presence of the yellow rubber duck.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a yellow rubber duck floating among a glass of water liquid, ice cubes. which matches the description description of the '(s The elements like the blue and ice cubes and blue liquid do not contradict the main of the yellow rubber duck.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4b5067d1b29ba881a61c4b3edaca9505779cc62b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3bb581600d993f0b517638cbee4f85197877f06ecdc28cadf970c823a1a2104
+size 81961
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..e9aa1dfaa042f01286b542f8212e2a70091e8297
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a woman with blonde hair and a striped shirt. She is sitting in a chair and appears to be engaged in a conversation with another person, who is not visible in the frame. The woman is wearing earrings and has a serious expression on her face. The background of the video is blurred, but it appears to be an indoor setting with a floral arrangement. The text overlay on the video reads \"FONDA & REDFORD REUNITED HOLLYWOOD LEGENDS PAIR UP FOR NEW FILM\" and \"MEGYN KELLY TODAY\". The style of the video is a standard news segment with a focus on the woman and the text overlay providing information about the topic being discussed."
+    ],
+    "video_ids": [
+        "KaAdS8Oso9U_8_127to329"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with blonde hair and a striped shirt sitting in a chair, wearing earrings and having a serious expression; another person not visible in the frame.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair, wearing a striped shirt, seated in a chair, with visible earrings and a serious expression. Another person is partially visible on the right side of the frame, but their face is not fully shown, fitting the description of 'another person not visible in the frame.' The core description is accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair and wearing a striped shirt, sitting in a chair, and earrings earrings. a serious expression. The person is not visible in the right side of the frame, but they face is not fully shown, and the description of 'another person not visible in the frame'. The overall elements is largely represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a0cd04ae007877dec9f9b9f1954b12705cfcec50
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1f444c1a6c2daab9e41c80c64274b782f52a7af54b0f04eb71b31b9da03c5206
+size 134218
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..cdd9cce3582c6580256abbc7ff4bdc27ef8f0203
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a vibrant yellow Audi TT car on display at a car show. The car is positioned in the center of the frame, drawing attention to its sleek design and shiny exterior. The car's grill and headlights are clearly visible, highlighting its modern and stylish features. The car show setting is indicated by the presence of other cars in the background, suggesting a variety of makes and models on display. The overall style of the video is dynamic and engaging, capturing the essence of the car show experience."
+    ],
+    "video_ids": [
+        "9ixgKv-U9tE_6_0to126"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A vibrant yellow Audi TT car.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a vibrant yellow Audi TT car, clearly visible from the front with its headlights on and the 'Audi TTS' badge displayed. The car is the central focus, and although other cars and exhibition elements are present in the background, they do not contradict the core description of the yellow Audi TT car.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a vibrant yellow Audi car car, which matching in the front. its distinctive and. the AudiAudi'TS' badge visible. The car is the central focus, and the there cars are a elements are present in the background, they do not detr the main description of the ' Audi TT..\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0b0c555669cc020095ef18424c04d71f1c4860b4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab7c2917966c62b432c0f76372c1f4e6b837ec129d989bbe1df3171f36d0cf51
+size 209541
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..d91eaeb77dea896166cb18cfb3e32fb013032a64
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man sitting in a chair, wearing a white shirt. He appears to be in a professional setting, possibly an office. The man is looking to the side, and his expression is serious. The background is simple and uncluttered, with a plain wall and a framed picture hanging on it. The lighting in the room is soft and even, suggesting an indoor environment. The man's posture is relaxed, but his attention seems to be focused on something outside the frame of the video. The overall style of the video is straightforward and unembellished, with a focus on the man and his surroundings."
+    ],
+    "video_ids": [
+        "1EuzdVJlzIw_86_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man sitting in a chair wearing a white shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting in a chair, wearing a white shirt with a checkered pattern, which matches the description. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting in a chair wearing and a white shirt, a collarered pattern. which align the description. The setting includes the elements in not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1cce8a8a004727f11236fd1fd9823ab59f37df63
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3ffd6e46ca7cc3fe3e63e227f54a4e886d9c4167ad4ebea1cf28c7d9e5a2ec64
+size 70378
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..afd39bc59d8a7500b5b394737c1df5105f44f2e5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment on a basketball court. A player, donned in a blue and red jersey with the number 21, is the focal point of the scene. He is in motion, running towards the camera with a triumphant smile on his face. His right hand is raised, making a peace sign, while his left hand is raised, holding a black mask. The background is a blur of spectators, their faces indistinct, adding to the sense of speed and movement. The player's jersey, the peace sign, and the mask are the main objects in the video, creating a narrative of a victorious moment in a basketball game."
+    ],
+    "video_ids": [
+        "QB9gwI52c4E_4_0to133"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Player, blue and red jersey with number 21, peace sign, black mask? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a basketball player wearing a blue and red jersey with the number 21, making a peace sign with his hand, and wearing a black mask. All elements described in the condition are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a basketball player wearing a blue and red jersey with the number 21. which a peace sign with one hand, and wearing a black mask. The these in in the ' are present in match depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a93a8805dfe61e470612e7ad6f81fc055a854bc5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:19472fd328e91fd1908951296f12a841371021e1bedfd424b58fce3490deb9d7
+size 229126
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..158eb6570b71513526de2ae91658e2c06873556a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is a quarterback, dressed in a vibrant red and white uniform with the number 15 prominently displayed. He is in the midst of a powerful throw, his arm extended and the football gripped firmly in his hand. His helmet, matching the color of his uniform, is a striking red. The background is a blur of activity, with other players and the crowd in the stands, creating a sense of anticipation and excitement. The style of the video is action-packed and energetic, capturing the intensity of the sport."
+    ],
+    "video_ids": [
+        "LxiAjXhxbzA_141_16to152"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Quarterback, football, helmet, other players, crowd in the stands? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a quarterback (wearing jersey #15) holding a football, wearing a helmet, and with a blurred background that suggests other players and a crowd in the stands. The elements are consistent with the description, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a quarterback inwearing a number15), in a football, wearing a helmet, and being other crowd background of includes the players and a crowd in the stands. The presence in consistent with a description of and the additional additional are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9f6c643613edb026c92195f151febcce33a09866
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d9aa4f4749e7453ed9e355879c2bdde8768ddac274fc04250ea55362f00d0d16
+size 268701
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..aad0476bc695851114f7c7e9084984d9fad52d6f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a vibrant yellow muscle car parked in a lot next to a black truck. The car is sleek and shiny, with a black hood and black rims that contrast with its bright color. The truck is large and imposing, with a black grill and black rims that match its color. The car is parked in front of a gray building with a white door, and there's a sign on the building that reads \"8100\". The car is the main focus of the video, with its bright color and shiny finish drawing the viewer's attention. The truck is parked behind the car, adding depth to the scene. The gray building and white door provide a neutral backdrop that allows the car and truck to stand out. The sign on the building adds a touch of intrigue, leaving the viewer wondering what it might represent. Overall, the video captures a moment of stillness in a parking lot, with the muscle car and truck as the main subjects."
+    ],
+    "video_ids": [
+        "HfLtypFVIyk_30_118to267"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A vibrant yellow muscle car and a large black truck.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a vibrant yellow muscle car (a Dodge Challenger) and a large black truck parked beside it, matching the core description. Additional elements like the building and lighting do not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a vibrant yellow muscle car,likely Chevrolet Charger) and a large black truck, in it. which the description description. The elements like the garage and the do not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d76ef907090e913595810a602b9419cf805dadd1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1239-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f218022a170fd6da5588ff54b27d0cf3ccfb400f62175318419e27b72cbd9a1d
+size 203487
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..3bbff0da6723e2afd21833511bf2ad3d2e6c2e89
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a kitchen, preparing a meal. She is holding a plate with a piece of orange on it. The kitchen is well-equipped with a refrigerator, microwave, and oven. The woman is standing in front of a counter, which has a bowl on it. The counter is made of granite. The woman is wearing a black shirt. The kitchen has a modern design with stainless steel appliances. The woman appears to be in the process of preparing a meal, possibly for herself or someone else. The overall style of the video is casual and homey, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "DFzBoiuE358_13_79to226"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Woman, plate with an orange, refrigerator, microwave, oven, bowl, black shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a black shirt holding an orange plate, with a refrigerator, microwave, oven, and bowl visible in the kitchen background. All specified objects are present and correctly identified, with no contradictions in the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a kitchen shirt standing a orange on. standing a refrigerator, microwave, and, and a visible in the background setting. The the objects are present and match identified.\"\n fulfilling no contradictions or the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..333f92990ddf39d97216ab13c9b083136f792e37
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30e3d1a5b02c8b93eae31d523836e4a606ac421076536a54b84608b39d566dd2
+size 152071
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..beb73543a378d5d759220a922521de9d8841e9ae
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, three people are gathered in a kitchen, engaging in a cooking activity. The kitchen is well-equipped with wooden cabinets, a sink, a stove, and a microwave. The counter is adorned with various cooking utensils and ingredients, including a plate of food, a bowl, and a cup. The people are standing around the counter, with one person actively cooking on the stove, while the other two are observing and possibly discussing the cooking process. The overall atmosphere of the video suggests a friendly and collaborative cooking session."
+    ],
+    "video_ids": [
+        "QAIDeOcqbGI_48_0to197"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three people, a plate of food, a bowl, a cup, a stove, a sink, wooden cabinets, a microwave, and cooking utensils.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three people standing in a kitchen setting. On the counter, there is a plate of food (pancakes), a bowl (with whipped cream), and a cup (with syrup). A stove is visible with pans on it, and a sink is in the background. Wooden cabinets, a microwave, and cooking utensils are also present, matching the described elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts three people in in a kitchen.. There the counter, there is a plate of food,possiblycakes), a bowl,green a cream), a a cup (with a). The stove and visible in a on it, and a sink is seen the background. The cabinets and a microwave, and various utensils ( also present. fulfilling the description elements.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..244f06d7f16b63b9015a356671231e78574bd032
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:560acdd36f771be59dedb9399a4515dacf0c7e91038d9c7605fca2ba1660005f
+size 165764
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..e28d2a256f84253654b3cd44cc0896449f22eff1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a car driving on a snowy road. The car is a dark-colored SUV with a visible spare tire on the back. The road is covered in snow, and the car's tires leave tracks in the snow as it moves. The surrounding environment is a mix of snow-covered ground and trees, suggesting a winter setting. The car's headlights are on, illuminating the snowy road ahead. The style of the video is realistic, with a focus on the car and its movement through the snowy landscape. The camera angle is from the side of the car, providing a clear view of the vehicle and its surroundings."
+    ],
+    "video_ids": [
+        "d9hRz3q7hIU_63_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A dark-colored SUV with a spare tire, snow-covered ground, and trees.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dark-colored SUV driving on snow-covered ground, with trees visible in the background. The SUV's tires are clearly visible, and the snow-covered terrain and surrounding trees match the description. Although no spare tire is visibly attached to the vehicle, the core elements described are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a dark-colored SUV with on a-covered ground with which trees visible in the background. The vehicle has spare are partially visible, and there ground on ground is trees trees match the description. The the spare tire is visible mounted to the rear, the presence elements of in present,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..12d4e17180f7c542a17c904cfeaebb933c8ff179
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:422808b965427a180faadad7738feb02913499137644d9c1f03acfccea21a09a
+size 332013
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..0a3ad5f2db14ac6705b1d999471cfb0d90b6d780
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a robotic kangaroo in motion against a dark background. The robot is predominantly white with black accents on its limbs and tail. It features a sleek design with a streamlined body and a long, flexible tail that aids in balance and movement. The kangaroo's legs are equipped with mechanical joints, allowing for fluid motion. The robot's body bears the word \"RESTO\" in blue letters, indicating it might be part of a series or brand. As the robot moves, its head turns slightly, giving the impression of awareness and responsiveness to its environment. The lighting is focused on the robot, highlighting its details and creating a stark contrast with the dark background. A small, glowing blue box appears in the bottom right corner of the frame, displaying the text \"15 lbs,\" which likely refers to the weight of the robot or an object it is carrying. The robot's movements are smooth and deliberate, suggesting advanced programming and engineering. The overall scene conveys a sense of innovation and technological"
+    ],
+    "video_ids": [
+        "ad127b14c0978d6aa47a425c524e8737dbc0f4492d3e8fde8b56da2bdf0e0476"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Robotic kangaroo, small, glowing blue box? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a robotic kangaroo with visible mechanical joints and the 'FESTO' logo, which matches the description. Additionally, a small glowing blue box displaying '15 lbs' is present in the lower right corner throughout the video, fulfilling the second part of the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a robotic kangaroo, a branding components and a textHSSTO' branding, which align the ' of Additionally, there small, blue box is the11'' is present in the bottom right corner, the video, fulfilling the ' part of the '.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f5bbe493be053255bf83a4b6d1119c6ee41adb14
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0eafdb8bab36bfc0879a74964dc646665e825c18ee1bd9d6aa0dee44eda2cbbf
+size 195515
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..7131ba43fd265c82ccfb51bde1a7ce06a8876788
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a red muscle car with orange rims, driving on a road. The car is sleek and shiny, with a classic design. The road is surrounded by a grassy field, and the sky is clear and blue. The car is moving forward, and the camera follows it from a side angle. The car's speed and the smooth motion of the wheels suggest a sense of power and freedom. The overall style of the video is dynamic and energetic, capturing the essence of the car's performance and the beauty of the open road."
+    ],
+    "video_ids": [
+        "GtYBqerPFyc_7_28to178"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red muscle car with orange rims.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red muscle car with orange rims, matching the core description. The car's vibrant red paint and orange-colored wheel rims are clearly visible throughout the frames. While there are additional elements like a black rear wing and background scenery, they do not contradict the primary description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a red muscle car with orange rims, which the description description. The car is vibrant red color and the wheels rims rims are clearly visible, the frames. The the is some elements like the blurred roof spoiler and a scenery, these do not contradict the main description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fb76aa03d22df369fe449763cf889170daf90a2b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:44ba68374bd3099d392f42946de1f565c7dc7e95caf693676e30b3120977abb9
+size 182789
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..e7fca3b30c5862142201c23cb1e4ef6884c82685
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man with white hair and glasses is seen working in a well-equipped workshop. He is wearing a black t-shirt and is focused on a task at hand. The workshop is filled with various tools and equipment, including a drill press, a workbench, and a shelf filled with different items. The man is seen using a drill to work on a piece of metal, carefully examining it under a magnifying glass. The overall atmosphere of the video suggests a sense of precision and craftsmanship."
+    ],
+    "video_ids": [
+        "Dt6BQPNGGTw_84_94to260"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with white hair and glasses, wearing a black t-shirt, and a drill press, a workbench, a shelf.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with white hair and glasses, wearing a black t-shirt, standing in front of a drill press on a workbench. In the background, there is a shelf with various items. All core elements described are present and consistent with the video content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man with white hair and glasses, wearing a black t-shirt, working at a of a drill press and a workbench. The the background, there is a shelf with various tools, The these elements of in present and match with the video content.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9e4129994c1b36dc695fa726f8fc06a4972ee368
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2f575922ba9eb8ea81384ec65267fb0e7db5e8b48b19808d803fd886edea263
+size 182061
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..b58d2861643672d1e14034d235aade08798029dd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is preparing a meal in a kitchen. The person is wearing gloves and is seen placing a piece of broccoli onto a cooling rack. The kitchen counter is cluttered with various ingredients, including a bowl of greens, a glass of wine, and a vase. The oven is visible in the background, suggesting that the meal is being cooked. The overall style of the video is casual and homey, capturing the everyday process of meal preparation."
+    ],
+    "video_ids": [
+        "an8SQPyOnQY_8_0to145"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person, broccoli, cooling rack, bowl of greens, glass of wine, vase, oven? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person wearing gloves placing greens onto a cooling rack, with a bowl of greens nearby, a glass of wine, a vase, and an oven visible in the background. All the specified objects are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person wearing gloves, broccoli onto a cooling rack. which a bowl of greens,. a glass of wine, and vase, and an oven in in the background. The the objects objects are present and the with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e366b9602a6ed441ba3218ff9adcb81725829807
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e02f003287d85ed12a2b94bd148108a75c505694baa8f90f836cdced921b7fd6
+size 154154
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..00e248faad440be788875f4adb8ad84f4bf4ea60
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man in a pink shirt and black apron is preparing a sandwich in a kitchen. He is standing at a counter with various ingredients and utensils around him. The kitchen has wooden cabinets and a window with blinds. The man starts by placing bread on the counter, then adds lettuce and tomato. He uses a knife to slice cheese and ham, and arranges them on the bread. He also adds condiments to the sandwich. The man is focused on his task, and the kitchen is well-lit, creating a warm and inviting atmosphere."
+    ],
+    "video_ids": [
+        "B82CFhJMdt4_8_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a pink shirt and black apron, bread, lettuce, tomato, cheese, ham, condiments, a knife, various ingredients and utensils.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a pink shirt and black apron, preparing a sandwich with bread, lettuce, tomato, cheese, ham, and condiments. He is also using a knife and various ingredients and utensils are visible on the counter. All elements described in the 'Object(s)' condition are present and consistent with the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a pink shirt and a apron, and a sandwich. bread, lettuce, tomato, cheese, and, and aiments. The is using using a knife to various utens and utensils, visible on the counter. The these of in the conditionObject(s)' condition are present and accurately with the video depicted}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..db091c90242aee623486997f0500409750fddedd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-124-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba3bb1a0aaaeb2b5f50e17e32654eedd16f119525c416fa584e825d120dba1f5
+size 124487
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..90ac33aa477a290784424fe1238c1d5bc98ca5d5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man singing passionately into a microphone. He is dressed in a dark blue t-shirt and a black baseball cap. The background is a blur of blue and white, suggesting an outdoor setting. The man's stance and the microphone suggest that he is performing in front of an audience. The overall style of the video is dynamic and energetic, capturing the essence of a live performance."
+    ],
+    "video_ids": [
+        "Tp8UmTL-DAI_18_0to158"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man dressed in a dark blue t-shirt and a black baseball cap, standing in front of a microphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a dark blue t-shirt and a black baseball cap, holding a microphone and appearing to perform. These elements match the core description provided. Additional elements like stage lighting and another person in the background do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a dark blue t-shirt and a black baseball cap, holding a microphone close singing to sing or The elements match the description description provided. The elements such the lighting and a person in the background are not contradict the main and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b6c97fb43b535c38b875d6ab9aea25885b4ce46c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0ddfcab6ec659458938efcce54cb7cf0c2c4c290a059546da79a29bdab03f69b
+size 173966
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..9e14853dd49946f1729caefefa2adb59a76331f9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man walking along a stone wall, possibly a part of a historical site or landmark. He is dressed in a black jacket and sunglasses, and he carries a backpack. As he walks, he points to something off-camera, drawing attention to it. The wall is made of large, rectangular stones, and it stretches into the distance. The man is the main subject of the video, and his actions suggest that he is either explaining something or showing something of interest to the viewer. The style of the video is straightforward and documentary-like, focusing on the man and his interaction with the environment."
+    ],
+    "video_ids": [
+        "5z2SjB5z9g0_29_20to160"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a stone wall, a backpack? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing sunglasses and a backpack, standing on a stone wall that is part of the Great Wall of China. The man is the central subject, and the stone wall and backpack are prominent and accurately depicted as requested in the 'Object(s)' condition. Additional elements like other tourists and the sky do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man wearing a and a dark, standing in a stone wall. appears made of a background Wall of China. The presence is pointing main figure, and the stone wall and backpack are visible features integral depicted, per. the descriptionObject(s)' condition.\"\n The elements like the stones or the clear are not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2d8b777b1d31d5cf48fb2bebe753cd43c13667cf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5850fa2093aac3f2655e4c9006ba60daf618d8f8fc23b04ac192a67fb29fca1
+size 201705
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..fade101d1ffbff3ca5b0640ce9e5e4059a81454d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man and a woman enjoying a day at the beach. The man, wearing a black cap and sunglasses, is seen taking a selfie with the woman. They are standing on a wooden deck that overlooks the ocean. The deck is surrounded by a stone wall, adding a rustic charm to the scene. In the background, the vast expanse of the ocean stretches out, with waves crashing against the shore. The sky above is a clear blue, dotted with a few clouds. The overall atmosphere of the video is relaxed and leisurely, capturing a perfect day at the beach."
+    ],
+    "video_ids": [
+        "7TA1rmGu2n8_44_0to180"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman as the main subjects, both visible in the foreground and interacting with the camera. Their presence is consistent throughout the frames, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man and a woman standing the main subjects. both wearing and the frame. wearing with each camera. They attire and central with the frames, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..87aa97a414542072dd9e2573e1de48db003bc691
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1752b04a9ee0bd26972ff31ca80a8a98f70ac8b0950c09726015dc52219c478c
+size 184406
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..51b7edeeeb5438dc6cbbcf818729e07434fd6f91
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two men are engaged in a conversation in a colorful, urban setting. The man on the left is holding a microphone, suggesting that he is the interviewer or host. The man on the right is holding a yellow microphone, indicating that he is the interviewee or guest. Both men are smiling and appear to be enjoying the conversation. The background features a vibrant cityscape with buildings and a skyline, adding to the lively atmosphere of the scene. The overall style of the video is casual and friendly, with a focus on the interaction between the two men."
+    ],
+    "video_ids": [
+        "ZAgXef21F3g_32_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, one holding a microphone and the other holding a yellow microphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men sitting at a table with microphones. The man on the left is holding a standard microphone, and the man on the right is holding a yellow microphone. This matches the 'Object(s)' condition described, even though there is also a skeleton in the background, which does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men, in a table, microphones. One man on the left is holding a microphone microphone, while the man on the right is holding a yellow microphone. The matches the descriptionObject(s)' condition described in as though the are an a microphone in the background, which is not conflict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6a2091aa2e4ecd3307b806bb15f84c5b5d1ce114
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd6fa8e9b924c9418dbb4f68729088a94fce8447a59b6ed846226fc832f9446d
+size 188120
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..c4aa9731d278bfd1d53e3d881df3a1a4da8827c5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is a quarterback, dressed in a vibrant red and white uniform with the number 15 prominently displayed. He is in the midst of a powerful throw, his arm extended and the football gripped firmly in his hand. His helmet, matching the color of his uniform, is a striking red. The background is a blur of activity, with other players and the crowd in the stands, creating a sense of anticipation and excitement. The style of the video is action-packed and energetic, capturing the intensity of the sport."
+    ],
+    "video_ids": [
+        "LxiAjXhxbzA_141_16to152"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Quarterback, football, helmet, other players, crowd in the stands? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a quarterback (wearing jersey #15) holding a football, wearing a helmet, and with a blurred background that suggests other players and a crowd in the stands. The elements are consistent with the description, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a quarterback inwearing a number16) in a football, wearing a helmet, and standing other crowd background that suggests the players and a crowd in the stands. The presence in consistent with the description of and the conflicting elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..49ea4392f03e861d0e5730c889baea0358703588
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a464845df1fea1645f312ede3074d284c0bc0cff320e668457ba6cf98de6707
+size 268619
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..d9dafbe4a3aa9e72e67991bcc94c40fff1967b67
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a red Ram truck parked in a parking lot under a highway overpass. The truck is positioned in the center of the frame, with the overpass towering above it. The truck is facing the camera, showcasing its front grille and headlights. The parking lot appears to be empty, with no other vehicles or people visible. The overpass is made of concrete and has multiple lanes, suggesting it's a busy highway. The sky is clear and blue, indicating good weather. The style of the video is realistic, with a focus on the truck and its surroundings. The camera angle is slightly elevated, providing a clear view of the truck and the overpass. The lighting is natural, with no artificial lighting visible. The video does not contain any text or additional graphics."
+    ],
+    "video_ids": [
+        "RiXVRmplp4s_44_108to277"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red Ram truck? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red Ram truck as the central subject, parked under an overpass. The truck is clearly visible and matches the description, with no conflicting elements that contradict the 'Red Ram truck' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a red Ram truck, the main object. which under a overpass. The truck's clearly visible with matches the description of fulfilling its conflicting elements present would the coreRed Ram truck' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cfa9e0059559e14eb71b0e041d6e058851c8bfee
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:903e166df375349898848846183a47859088fb321b68ee73ce6a7940710c1894
+size 120891
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..aafe4f44fffccc95867082fb81c2cf63868bf721
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a woman in a white chef's uniform, standing in a kitchen. She is stirring a pot with a spoon, and the pot is placed on a stove. The kitchen has white tiles on the walls and a stainless steel countertop. The woman is focused on her task, and the kitchen appears to be well-equipped and clean. The style of the video is realistic and it captures the woman's actions in a clear and detailed manner."
+    ],
+    "video_ids": [
+        "Z8LR0aGVvEY_18_0to102"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a white chef's uniform, a pot on a stove, and a spoon.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a white chef's uniform stirring a pot with a spoon, which matches the core description. The pot is on a stove, and the spoon is being used for stirring. Additional elements like bowls and kitchen equipment do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a white chef's uniform cooking a pot on a spoon on which align the description description. The presence is on a stove, and the woman is being used to cooking, The elements like the and a tiles in not contradict the core and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2e5e02c17099c4c8c4e42f7a2c38da2f74b13acc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:01809364b33960c83772fece31b8ab3281e442e1c13ba0cecadb2760f843f315
+size 104274
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..3eacaef0338b05277ddb26ce87f175291ebbf2a2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a red Volkswagen Atlas SUV, captured in three different frames. The first frame displays the front of the car, highlighting its sleek design and the Volkswagen logo. The second frame focuses on the side of the car, revealing its robust structure and the red color that dominates its exterior. The third frame provides a rear view of the car, emphasizing its overall design and the Volkswagen logo once again. The video style is straightforward, with a clear emphasis on the car's design and features. The car is stationary, and there are no additional elements or actions in the video. The lighting is consistent across all frames, ensuring that the car's details are clearly visible."
+    ],
+    "video_ids": [
+        "Kihu3zheq9E_21_0to140"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red Volkswagen Atlas SUV? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red Volkswagen Atlas SUV, clearly identifiable by the Volkswagen logo on the grille and the 'ATLAS' lettering on the front bumper. The vehicle's design and color match the description, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red SUV Atlas SUV from which matching by its Volkswagen logo and the front and the overallAtlasLAS' nameing on the rear.. The vehicle is design, features match the description of and there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..44c6dabdd0676cc60ab872885f4670be18c1d214
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1240-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41857c07f7a9e00314c1b4ff1ba742f27a1f1f171e783f96b61531bcd7a19415
+size 181780
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..12ea68bac86be9bc0103b79cb733a5d4b6194cfb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment on a soccer field where a group of players, dressed in red and blue uniforms, are gathered. The players are standing on the lush green field, their attention focused on something off-camera. The coach, distinguishable by his red jacket, stands out among the players. He is positioned in the center of the group, his hands casually tucked into his pockets as he observes the scene. The players are scattered around him, some closer to the camera and others further away, creating a sense of depth in the image. The overall atmosphere is one of anticipation and focus, as if the players are waiting for the coach's next instruction."
+    ],
+    "video_ids": [
+        "Q-ySjtSbBFY_5_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Players in red and blue uniforms, a coach in a red jacket? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple players wearing red and blue uniforms, and a coach in a red jacket standing prominently in the foreground. The attire and setting match the description, with no conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows players players wearing red and blue uniforms, which there person in a red jacket. among in the center. The scene and the are the description provided indicating no significant elements present}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..62db7ce21e28578da011f6358455c0844bcb1c52
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a6851b42b22e49f44a1ee0c87732e88ec7bc9818a5d210b8eabd311fc51a53d0
+size 159960
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..fcc335d63c809207eaf21cc30a59d2186b7721c4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of the interior of a car, focusing on the front seats. The seats are upholstered in a light gray or beige material with a diamond-stitched pattern. The headrests are adjustable and feature a logo on the top center. The seat belts are visible, with the buckles located on the side of the seats. The car's interior appears to be well-maintained and clean. The style of the video is straightforward and documentary, with no additional elements or embellishments. The focus is solely on the car's interior, providing a clear and detailed view of the seats and their features."
+    ],
+    "video_ids": [
+        "FpoScPVWH10_17_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Seats, headrests, seat belts, buckles, logo? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the interior of a car with two front seats, headrests, seat belts, buckles, and a visible logo on the headrests. All specified elements are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows seats interior of a car with seats seats seats, headrests, seat belts, andles, and a logo logo on the seatrest.. The the elements are present and match depicted,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9f81431ea8a201c1996499ccd28da72c43e1e55a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c97ea701d0215bcd274f60bb0f2ecfa9a39694ad76438f451670433f24c510be
+size 136672
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..534a294cf8ddb8017798ec255a31fa866d6f8257
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car from the perspective of the passenger seat. The car has a luxurious design with a brown leather interior. The dashboard is equipped with a touch screen display and various buttons and dials. The steering wheel is also made of leather and has a red accent. The car's door features a cup holder and a window control panel. The car is parked in a garage with a white wall in the background. The video is a still image sequence, capturing the car's interior in three different angles."
+    ],
+    "video_ids": [
+        "2C2X7rEdbE4_7_0to186"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: touch screen display, buttons and dials, steering wheel, leather door, cup holder, window control panel, white wall? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a touch screen display, buttons and dials, steering wheel, leather seats (not explicitly a 'leather door' but likely referring to leather interior trim), cup holder, and window control panel. The white wall is visible outside the car through the window, matching the description. No elements contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful the car screen display, buttons and dials, a wheel, leather door,which door mentioned doordoorather door'), but part part to the seats),), cup holder, and window control panel. The white wall is not in the car window the window, fulfilling the '.\"\n The additional contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9c507565e135e74815be8c4da26d30b83e9f1c02
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:19e3493d71b05ec2a3ef5ec112aa9b25b6cf520a4fee41035752712bc8eee6c4
+size 177983
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..c5fc0be5fc4fa52b6ab810272b6153208f1ad418
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two men are engaged in a conversation in a kitchen. The man on the left, wearing a black chef's coat, is gesturing with his hand, possibly explaining a cooking technique or recipe. The man on the right, wearing a white apron, is attentively listening. They are standing in front of a wooden countertop that is adorned with various kitchen utensils and ingredients, including a bowl, a spoon, and a bottle. The kitchen is well-equipped with modern appliances such as a refrigerator and an oven. The overall atmosphere of the video suggests a friendly and informative cooking session."
+    ],
+    "video_ids": [
+        "Q3ryhVDB5iE_1_0to122"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, one wearing a black chef's coat and the other wearing a white apron.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men standing in a kitchen. One is wearing a black chef's coat, and the other is wearing a white apron with a logo. This matches the core description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two men in in a kitchen. One man wearing a black chef's coat, and the other is wearing a white apron. a chef on The matches the description description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..463972444587dc53f919dcbe261bee122c56ba18
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:58fe9f2e5a25be8612e1acab8fe14b6be2be2bf0e4e630c978dae9b209aca5c5
+size 81192
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..d79132dbd3b04788c384f47aa736ef6650007cbf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a woman with curly hair and glasses, wearing a black and white shirt with a floral design. She is standing in front of a red and white geometric patterned wall. In the first frame, she is looking up and to the left, with her hands on her head. In the second frame, she is looking down and to the right, with her hands still on her head. In the third frame, she is looking up and to the left again, with her hands still on her head. The style of the video is casual and candid, capturing the woman in a moment of contemplation or thought."
+    ],
+    "video_ids": [
+        "GqBHjBDgLfY_5_17to169"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with curly hair and glasses, wearing a black and white shirt with a floral design.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The woman in the video has curly hair and glasses, and she is wearing a black and white shirt with a floral design. These features match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video in the video has curly hair and is, and she is wearing a black shirt white shirt with a floral design. The elements match the description provided. the questionObject(s)' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4a1990e649df1004d6537845a374de5a966aa32d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75cfc43d5d5e2af9619eb93dd72db8767b1de268cd78ffa4b7fdd45e349cef87
+size 148188
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..2d338f9d5d134c2df76a68644abcd5f6a593011c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen holding two packets of snacks, possibly chips, in his hands. He is standing in front of a jeep, which is parked in a wooded area. The man is wearing a blue t-shirt and a baseball cap. He appears to be excited about the snacks, as he is holding them up to the camera. The jeep is parked on a dirt road, and the surrounding area is filled with trees and foliage. The man's excitement and the outdoor setting suggest that this could be a road trip or an adventure."
+    ],
+    "video_ids": [
+        "bwuwLD6xYck_3_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, two packets of snacks (possibly chips), a jeep? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man holding two packets of snacks (labeled as 'Eel-Oh! Snack Pack'), standing in front of a jeep. All core elements described are present and consistent with the prompt, with no contradictory elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man holding two packets of snacks,whichabeled as 'KWE'Style!'),acks')'), and in front of a jeep. The elements elements of in present: match with the video. making no additional elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..57d486c2dcaaa16aa7574cc76c21602606b93771
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac620c47ba819e55855ad271bf9a015e3434605342bceb0c60ef5d38a22de170
+size 364586
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..29971140c27cc487f72da85b223c50ddab8ed838
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man's journey along a rocky beach. He is dressed casually in a black t-shirt and sunglasses, carrying a backpack. The man is seen walking towards the camera, his gaze fixed on the distant horizon where a mountain range meets the ocean. The beach is dotted with a few other people, some of whom are enjoying the water. The sky above is clear, suggesting a sunny day. The overall style of the video is a simple, unobtrusive documentary, capturing the man's journey and the natural beauty of the beach and ocean."
+    ],
+    "video_ids": [
+        "ck0HqHk1RH4_18_0to212"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black t-shirt and sunglasses, carrying a backpack; a few other people on the beach.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a man wearing a black t-shirt and sunglasses, with a backpack visible on his shoulders, which matches the core description. Additionally, there are other people on the beach, including individuals near the water and under umbrellas, which aligns with the secondary condition. The scene is consistent with a beach setting, and no elements contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a man wearing a black t-shirt and sunglasses, carrying a backpack, on his back. which align the ' description. Additionally, there are a people visible the beach in which one in the water and others therellas, which aligns with the description part. The setting is set with the beach setting, and the elements contradict the given.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c3c46a475bb68b882ae6c29fe120c5a7356988a3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3e671ca58ab44e5ce9c576f941aec65c72fd1e9376612432029c6a6c7e0f406
+size 195979
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..11bf69c74e621cfa855ff010c255c3ba3ab20b16
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a lively cooking show featuring three men in a kitchen. The men are standing around a large metal table, which is filled with various ingredients and cooking utensils. The man on the left is wearing a hat and is pouring a drink into a glass. The man in the middle is holding a glass of tea and is looking at the camera. The man on the right is holding a knife and is cutting a lemon. The kitchen is well-lit and has a modern design. The men are all smiling and seem to be enjoying themselves. The video is shot in a documentary style, with a focus on the men and their actions."
+    ],
+    "video_ids": [
+        "Je4Wwm3-Y8g_45_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three men, a large metal table, various ingredients, cooking utensils, a glass, a glass of tea, a knife, a lemon.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three men standing around a large metal table, mixing drinks with various ingredients and utensils. There are glasses, bottles, and citrus fruits (including lemons) visible. While there is no explicit glass of tea or knife shown, the core elements described \u2014 three men, a large metal table, various ingredients, and cooking utensils \u2014 are all present and consistent with the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts three men in around a large metal table. which ingredients and various ingredients and utensils. The are bowls, a, and a fruits likele lemons) on, The the are no knife mention of tea, knife,, the presence elements of in the men, a large metal table, various ingredients, and cooking utensils \u2014 are present present and match with the description depicted}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8baaf6dfdd4c733642f489b518979c05d0fbe97f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1241-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:690fa3b7f6f848f9d5c556be025a607333bef9c544398546c5ca2a5f31005982
+size 137393
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..81d7856966a786eec8afbdbd590481594aaae8a1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a red car in motion on a race track. The car is sleek and modern, with a shiny finish that reflects the sunlight. The track is wide and well-maintained, with clear markings and a smooth surface. The car is moving at a high speed, as indicated by the blurred background and the slight tilt of the car's body. The car's wheels are firmly planted on the track, and the car's body is angled slightly to the left, suggesting a turn. The car's headlights are on, and the car's taillights are visible, indicating that the car is moving forward. The car's license plate is clearly visible, and the car's brand is identifiable. The car's design is aerodynamic, with a low and wide stance. The car's windows are tinted, and the car's doors are closed. The car's roof is smooth and unobstructed. The car's wheels are black and shiny, with a multi-spoke design. The car's tires are wide and low-profile, suitable for high-speed driving. The car's suspension is low, and the car's body is slightly lowered, indicating a sporty design. The car's exhaust pipes are visible, and the car's engine is likely powerful. The car's interior is not visible in the video. The car's exterior is red, and the"
+    ],
+    "video_ids": [
+        "JVntSkOHPQE_2_66to246"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red, sleek, and modern car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red car that appears sleek and modern in design, consistent with the description. The car's styling, including its aerodynamic lines and contemporary features, matches the 'sleek and modern' criteria. While the background includes a racetrack and distant scenery, these elements do not contradict the core description of the car itself.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a red car with appears sleek and modern, design. consistent with the description. The car's design, including its aerodynamic shape and sport features, align the 'sleek and modern' aspect. The the video includes a raceetrack, green trees, these elements do not contradict the core description of the car itself.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3a2be1f8a4df5359cc517479676c514467a7da97
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a2015b87d6ba8bbebcfcd15e4ad8fc1abf5064bd7bc84de0f2c99619514b63ac
+size 229326
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..d143040af4875084f0ef5e8b44f80908fc9b744a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and glasses, wearing a green hoodie and a black t-shirt with a colorful graphic on it. He is standing in front of a wooden fence and a tree. The man appears to be speaking or presenting, as suggested by the text overlay on the video. The text reads, \"This information isn't specific to NordPass, but rather the security people behind it (who also built NordVPN).\" The style of the video seems to be informative or educational, possibly related to cybersecurity or online privacy. The man's attire and the casual setting suggest a relaxed and approachable tone."
+    ],
+    "video_ids": [
+        "BKSSxWBOTN8_73_149to322"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and glasses, wearing a green hoodie and a black t-shirt with a graphic.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The man in the video clearly has a beard and glasses, is wearing a green hoodie over a black t-shirt with a visible graphic, and these features are consistent throughout the frames. The background elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video in the video has has a beard and is, and wearing a green hoodie, a black t-shirt with a colorful graphic. which the elements match consistent with the frames provided The presence and, not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8e792e07511595b594ef7220d2e09bfba067e8d1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ef9c35c200de077983eaa1a348fda973ed45d932e2a11003d32c8ee7d13808f
+size 156615
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..e0cb791997ae98bc96647605bb9e548ebb9d8d0b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a kitchen, holding a bowl filled with white foamy substance. She is wearing a blue shirt and glasses, and her facial expression suggests surprise or confusion. The kitchen is well-equipped, with a refrigerator, oven, and sink visible in the background. There are also various items on the counter, including a bottle, a vase, and a potted plant. The woman's hand gestures and the context of the scene suggest that she is either preparing a recipe or reacting to an unexpected outcome. The overall style of the video is casual and candid, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "XoWgMy6IOsE_7_61to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a bowl filled with white foamy substance, a blue shirt, glasses, a bottle, a vase, and a potted plant.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing glasses and a blue shirt, holding a bowl filled with a white foamy substance. In the background, a bottle, a vase, and a potted plant are visible, matching the described elements. All core objects are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a and a blue shirt, holding a bowl filled with a white foamy substance. There the background, there bottle, a vase, and a potted plant are visible, which the description elements. The the conditions and present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f66f6d27cab498ac099f1f588f3210dd1773eb6c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0eb0c0ca2a1425fb2c97908a4b36995fb44b60d8ca406434dfd26a99157990a1
+size 172746
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..f51efaccec1050ca1d17a6472844c37627effa25
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. A player, wearing a white jersey with the number 11, is in the process of catching a red football. He is running towards the right side of the frame, with his arms outstretched to secure the ball. His focus is intense, and his body language suggests a high level of skill and experience.  In the background, other players can be seen, some in pursuit of the ball, while others are in various positions on the field. The field itself is a vibrant green, contrasting with the players' colorful uniforms. The atmosphere is charged with anticipation, as the outcome of the play hangs in the balance.  The style of the video is realistic, with a focus on capturing the action and intensity of the game. The camera angle is dynamic, following the player as he moves across the field, and the lighting is natural, suggesting that the game is taking place during the day. The overall effect is a thrilling snapshot of a moment in a football game, filled with potential and excitement."
+    ],
+    "video_ids": [
+        "6iaApxa661g_17_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a white jersey with number 11, a red football, other players in pursuit.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a white jersey with the number 11, holding a red football, and other players are visible in pursuit in the background. These elements align with the described 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a player in a white jersey with the number 11, holding a red football. and there players in in in pursuit. the background. The elements match with the description 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d3a73274f19888a888fdfafeba51bae8e0dfbf90
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d1d1e7a26bc9e2e1112e051d059f726295235197c4731cdd094b9faf75afc6da
+size 273978
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..4d4f7dd3f139ef96bc6008deba17270d7941741e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a mesmerizing underwater scene where a large school of fish swims in a circular pattern, creating a tunnel-like effect. The fish, silhouetted against the bright sunlight filtering through the water's surface, move gracefully in unison, their bodies forming a continuous line that spirals outward from the center of the frame. The water is a deep blue, with gentle ripples and light reflections adding texture to the scene. As the video progresses, the fish maintain their synchronized movement, their shadows elongating and shortening as they navigate the circular path. The sunlight creates a radiant glow at the top of the frame, enhancing the ethereal quality of the underwater environment. The overall atmosphere is serene and captivating, highlighting the beauty and harmony of marine life."
+    ],
+    "video_ids": [
+        "39b023d1a1dc03e53254afe8080b3037389fb7ac31025b06d5c6b8a0e2d86b7b"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large school of fish? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a large school of fish swimming in a circular pattern underwater, with sunlight filtering through the water above. This matches the description of 'A large school of fish' without any conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a large school of fish swimming in a coordinated pattern,. which the filtering through the water above. The matches the description of aA large school of fish' as any conflicting elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d490b6e5209acc0a12cd64ccccbe37849ca4354d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f6c9d9fee0c7b450820b8e715b879b8a57897818bb3bb81e370feb7b4739531
+size 339599
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..c095b1fae8d1c0460b23dfabd8d659c5bcb64051
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a blue shirt sitting under a tree. He appears to be in a relaxed state, with his eyes closed and mouth slightly open. The man is leaning against the tree, suggesting a casual and comfortable setting. The background is blurred, but it seems to be an outdoor environment with natural light. The style of the video is candid and informal, capturing a moment of rest or contemplation."
+    ],
+    "video_ids": [
+        "AMkyMNlSHqw_2_0to119"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a blue shirt, sitting under a tree, leaning against it? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue shirt sitting on the grass under a tree, with his body leaning against the tree trunk. This matches the core description provided. The presence of a woman sitting nearby does not contradict the description, as the focus is on the man and his position relative to the tree.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue shirt, under the ground under a tree, leaning his back leaning against the tree trunk. The matches the description description of, The presence of the tree in on in not contradict the main as as it main is on the man in his position under to the tree.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..302bd80f356aa366baba3b595301248e5e54716f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:33cf633af5fa023b8e3896e31d8f8b54fe050c594774ec4c99fee9c39e634dab
+size 126830
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..f423dd208501df6cee1af0738bf8a22416a86b9c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a pile of fresh green spinach leaves on a wooden surface. The leaves are vibrant and appear to be freshly washed, with some leaves overlapping each other. The focus is on the texture and color of the spinach, highlighting its freshness and natural beauty. The style of the video is simple and straightforward, with no additional elements or distractions. The lighting is bright and even, enhancing the natural colors of the spinach. The video does not contain any text or additional objects. The overall impression is one of simplicity and natural beauty, focusing on the freshness and appeal of the spinach leaves."
+    ],
+    "video_ids": [
+        "vsGMYPzpDqs_4_55to209"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A pile of fresh green spinach leaves? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by showing a clear, close-up view of a pile of fresh green spinach leaves. The leaves are vibrant, appear fresh, and are arranged naturally on a wooden surface, matching the description accurately. The only other element in the video is a brief clip of a man in a suit, which does not contradict the core description of the spinach leaves.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as depicting a pile and close-up view of a pile of fresh green leaves leaves. The leaves are vibrant green and to, and are arranged in, a wooden surface, which the description of.\"\n There lighting additional element present the video is the wooden glimpse of a wooden speaking a suit, which does not conflict the main description of the spinach leaves.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6d56c382ddc107ae4f81d83b3398bb7d66f00721
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e39e070eeeddf5b7f4c3c8d6fad439c2c43c3d5e6e8511c13cee434759ed2f8a
+size 69517
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..d43ad500934d82e945b2dd03d6e71051a3455aa7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man working on a car engine in a garage. He is wearing a gray t-shirt and glasses. The garage is equipped with various tools and equipment, including a toolbox and a workbench. The man is using a wrench to work on the engine, which is open and exposed. The car is a modern model, and the engine is complex with many parts visible. The man appears to be focused on his task, and the garage is well-lit, indicating that it is daytime. The style of the video is a straightforward, real-life depiction of a mechanic at work."
+    ],
+    "video_ids": [
+        "5G9bDynhVpc_5_617to766"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a modern car with an open engine, a toolbox, a workbench, and a wrench.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man working on a modern car with its engine open, and the background includes a toolbox and workbench. Although a wrench is not clearly visible in the frame, the man is using tools, and the context strongly implies the presence of a wrench. The core elements described are all present and consistent with the video content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man working on a modern car with an engine open. which he presence includes a toolbox and abench. The the wrench is not explicitly visible, the video, the man is using a, which the context suggests suggests the presence of a wrench, The video elements of in present present, match with the video content.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..deab15495b59806fdb4f6b4fe56aefea730713d0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1242-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:00f467e2ac2d461f27b26dd2acff71aff12e189a7f931e518da58e36646a2e99
+size 253489
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..e6253fa45c38a77153a1402d614c013284ffdd9a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a car's dashboard and center console. The dashboard features a digital display showing the car's battery level, fuel level, and other information. The center console has a touch screen display, which is being interacted with by a person's hand. The car's air conditioning vents are visible, and the car's interior is well-lit, suggesting it is daytime. The style of the video is a straightforward, real-life depiction of a car's interior, focusing on the technology and controls available to the driver."
+    ],
+    "video_ids": [
+        "Itb0C8_2-PA_9_816to962"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Digital display on the dashboard, touch screen display on the center console, a person's hand interacting with the touch screen, car's air conditioning vents.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a digital display on the dashboard (instrument cluster) and a touch screen display on the center console. A person's hand is visible interacting with the touch screen, and the car's air conditioning vents are prominently displayed below the screen. All elements described in the condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a digital display on the dashboard,top cluster), and a touch screen display on the center console. A person's hand is interacting interacting with the touch screen, and the car's air conditioning vents are also displayed and the touch. The elements described in the question are present and accurately depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..94c500cd3862e7f7c72247bae774779cecc24e92
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:27e9578a1190e77cd42d6c554ac6c2cacb7b537332df867899f99c6f07116370
+size 163894
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff61fa4edc2d82899f4fe94a8051ccfb27cc91e4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man working on a mechanical object, which appears to be a metal engine or machine part. The man is holding the object with both hands, and he is using a tool to work on it. The setting is a workshop or garage, with various tools and equipment visible in the background. The style of the video is a straightforward, real-life depiction of a person engaged in a manual task. The focus is on the man and his work, with no additional embellishments or artistic effects. The video captures the process of working on a mechanical object, highlighting the precision and skill involved in such tasks."
+    ],
+    "video_ids": [
+        "0QL8g0ums9c_38_52to204"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a mechanical object (metal engine or machine part), and a tool.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man working on a mechanical object, which appears to be a metal engine or machine part, and he is using his hands as tools to manipulate the components. The scene is consistent with the described conditions, with no significant contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person interacting with a mechanical object, which appears to be a metal engine or machine part. using using is using a hands as tools to adjust it object. The presence is set with the ' conditions, as the additional additional.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..44b3c322a0e8f7537468939e694e3322b14357c2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a40ccac5ec38f05b1d30cc2ed4d9bc213ea4576389903cd20386b4ea3714073
+size 139099
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..12429b2b82ee2d4c9a92710dae66fc2de315f2d9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit and tie, sitting in an office environment. He is gesturing with his right hand, pointing towards something off-camera. The office has a white wall in the background, adorned with a map of the United States. The man appears to be engaged in a conversation or presentation, as suggested by his expressive hand gesture. The overall style of the video is professional and formal, with a focus on the man and his interaction with the office environment."
+    ],
+    "video_ids": [
+        "HhYbDpfdukk_11_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, a map of the United States? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a suit and tie, and there is a map of the United States on the wall behind him. The man is the central focus and the map is visible in the background, fulfilling the 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man in a suit and tie, and there is a map of the United States in the wall behind him. The presence is gest main figure, is map is a in the background, fulfilling the 'Object(s)' condition as any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..faf06822edb9c3df455989de55bf5d0d91dd10f6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34503758f2358387185fde95395a950f91366c1246a6d793beb6b2cb628911dc
+size 187527
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..53059a939a9bad4cff2b649d7f73e16b212ecdc2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car, focusing on the front seats. The seats are gray and have a quilted design, with the letter \"R\" embossed on the headrests. The car's interior is well-lit, highlighting the seats and the dashboard. The dashboard features a sleek design with a touch screen display. The car's door is open, revealing the door handle and window controls. The overall style of the video is sleek and modern, showcasing the car's interior design and features."
+    ],
+    "video_ids": [
+        "QZBsvc0AAg8_58_0to157"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Seats, headrests, dashboard, touch screen display, door, door handle, window controls? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the seats and headrests, which are prominently featured and detailed. The dashboard is partially visible, including the gear shift and surrounding controls. The door and door handle are visible on the right side, and window controls are implied by the presence of the door panel controls. No elements contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows the interior, headrests of which are the featured. match. The dashboard and not visible, and the touch shift area part area. The touch is door handle are also on the right side of and the controls are implied by the presence of the window and.. The additional contradict the description,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..835677bde2fc79829de83818931a4018e9586281
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c522ba7e2b8ffd84d04dbe34e354256cb4353c801405186023d4f1abfef8a7c
+size 140240
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..56bc6e5009021ef74735765d47fb7013112f3c18
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the breathtaking view of a colorful canyon, showcasing its natural beauty and geological formations. The canyon is filled with a variety of rock formations, each exhibiting a unique color and texture. The colors range from warm hues of red and orange to cooler tones of blue and green, creating a vibrant and dynamic landscape. The canyon's geological formations are diverse, with some rocks appearing smooth and rounded, while others are jagged and angular. The video is taken from a high vantage point, providing a panoramic view of the canyon and its surroundings. The perspective allows for a comprehensive view of the canyon's depth and scale, emphasizing its grandeur and majesty. The video is a testament to the awe-inspiring power of nature and the beauty of geological formations."
+    ],
+    "video_ids": [
+        "7GX8tQ1bVTg_5_206to339"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Rock formations? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features colorful, eroded rock formations with distinct layers of yellow, pink, and white, which are the central focus. These formations are consistent with the 'Rock formations' object condition, and no elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a, layeredoded rock formations with various layers and different, orange, and blue, which are typical core elements. The formations are consistent with the 'Object formations' condition condition, as there other contradict this description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bb7d587223718803ef81c65c879814ef6908445c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a1820b4df38b184de41126f4f104aac0a3f3e5077d050cba3102f41e8b0a333
+size 249719
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..9dfda4973a00471479a6b647e29bbd0e93784f69
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man driving a black luxury car on a winding road. The car's interior is well-lit, with the dashboard displaying various controls and screens. The man is wearing a blue sweater and is focused on the road ahead. The car's steering wheel is visible, and the man's hands are on it. The road outside the car is curving, and the car is moving along it. The car's windows are down, and the man is enjoying the open air. The car's interior is spacious and comfortable, with leather seats and a well-designed dashboard. The man appears to be in control of the car, and the car is moving smoothly along the road. The video captures the man's journey in the luxury car, showcasing the car's interior and the man's driving experience."
+    ],
+    "video_ids": [
+        "eqHVjHTeHh0_8_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a black luxury car, a winding road, dashboard controls and screens, leather seats.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man driving a black luxury car on a winding road, with visible dashboard controls and screens, and leather seats. All elements described in the condition are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a man driving a black luxury car. a winding road. with visible dashboard controls and screens, and leather seats. The the described in the ' are present and match with the video content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e98cf36658a77f52fe407f0cea2a719bedc134ad
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:16cbaeefff655c11cbb813813f11603deb169f0424e0076a43c6b16908e858a7
+size 213478
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..e022d97dc0349b1d80e0a913fa5c20cef36ddffd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are seen riding in a golf cart. The man is driving the cart, while the woman is seated beside him. They are both wearing masks, indicating a concern for health and safety. The golf cart is moving through a lush, tropical garden, filled with vibrant greenery and colorful flowers. The garden appears to be well-maintained, with a variety of plants and trees. The man and woman seem to be enjoying their ride, taking in the beauty of their surroundings. The overall style of the video is casual and relaxed, capturing a moment of leisure in a beautiful setting."
+    ],
+    "video_ids": [
+        "CbXUGX8x4cw_205_80to209"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman seated in a golf cart. The man is driving and wearing a mask, while the woman is seated beside him, gesturing as she speaks. These two individuals are the central focus of the video, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man and a woman sitting in a golf cart. Both man is driving, the a mask, while the woman is seated beside him, alsouring with if speaks. The elements individuals fulfill the primary focus of the video, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..094122f881a8aa9a19f96490f084684c084d1042
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b7a82ee91a1a54092fea8aa3bd3a79f913947c51cc732ebf3d3a4ff5791f96b
+size 151165
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..2ce0827981bc64017908defc90cbd81ae45ecd04
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with a furry hood covering her head, looking down with a contemplative expression. The fur is light brown and appears soft. The woman is indoors, with a plain white wall in the background. There are no other people or objects in the scene. The lighting is soft and even, suggesting an indoor setting with ambient lighting. The style of the video is simple and straightforward, focusing on the woman and her expression."
+    ],
+    "video_ids": [
+        "5NSSGk2CwTI_29_0to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with a light brown, soft furry hood covering her head, looking down with a contemplative expression.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a light brown, soft furry hood that covers her head. She is looking down with a contemplative expression, which aligns with the description. The presence of hands adjusting the hood and the background elements (hanging caps) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a light brown, soft furry hood that covers her head. She is looking down with a contemplative expression, which aligns with the description provided The fur of the in the hood does the slight being dolikeanging clothes) do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1bd1812e2ed34cec9df44332dd0341fa25a95f6a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1243-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6696454d99c2888b6791105508d541cf1371f9ad5de8ebb00fc5a599cf4cdc64
+size 135737
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..e4de1818b12ecaab4cb516649e1d859485202bbb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up, high-resolution, and detailed view of a Yamaha dirt bike engine. The engine is silver and blue, with various parts and components visible. The style of the video is a time-lapse or a series of still images showing the engine in different stages or angles. The focus is on the intricate details and craftsmanship of the engine, highlighting its design and functionality. The video does not include any people or other objects, focusing solely on the engine. The lighting and composition of the video emphasize the metallic and mechanical aspects of the engine, showcasing its complexity and precision."
+    ],
+    "video_ids": [
+        "E0z3l2GD0UA_6_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Yamaha dirt bike engine? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a close-up of a Yamaha dirt bike engine, identifiable by the 'YAMAHA' branding on the engine cover and the typical design of a dirt bike engine with visible components like the exhaust pipe, clutch lever, and foot pegs. The blue and silver color scheme is consistent with Yamaha's branding for dirt bikes. There are no elements that contradict the description of a Yamaha dirt bike engine.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a close-up of a Yamaha dirt bike engine, which by the visibleYamahaHA' branding and the engine cover and the overall design elements a dirt bike engine. visible components like the crank,, engine,, and part pegs. The presence and black color scheme is also with Yamaha's design and their bikes. The are no elements in contradict the description of a Yamaha dirt bike engine.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..db6d1ad0d5ffe15a0a77ad3583a371cfe6b59d09
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:49091b24b7012f9f59371fc82df69ad1c13133bd8b5afc535e3b803299d98944
+size 152195
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..dd8eb99fe5627e832ca80a7e7a85cbc8f5bdd62c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman with short red hair is standing in a kitchen. She is wearing a blue shirt and has a surprised expression on her face. The kitchen is well-lit and features a white tile backsplash. There are various objects in the kitchen, including a sink, a window, and a shelf with various items on it. The woman appears to be in the middle of a conversation or reacting to something. The overall style of the video is casual and candid, capturing a moment in the woman's day."
+    ],
+    "video_ids": [
+        "6XPYz84JjW4_29_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with short red hair, a sink, a window, a shelf with various items.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with short red hair, standing in a kitchen setting that includes a sink and a window. Behind her, there is a shelf with various items, such as jars and decorative objects. All elements described in the condition are present in the video, and no conflicting elements are observed.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with short red hair, standing in a kitchen setting. includes a sink and a window. There her, there is a shelf with various items, which as containers and a objects. The these in in the condition are present in the video, making there additional elements are introduced.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7967c38915c958b966caa4e1264d85f680b64fb3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e9474f4f42d709dc44ad55d7ed0f5b1d12875f35b3d2d49e32138d4f9c346a20
+size 108005
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..cf6267e777bc4bdb42ace966966a598dfd677fa3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic scene at a race track. The first frame shows a red car speeding down the track, its tires hugging the curve of the track. The second frame shows the car continuing its journey, now rounding a bend in the track. The third frame shows the car approaching a green grassy area, its speed evident in the blurred background. The track itself is surrounded by lush green grass and trees, providing a stark contrast to the red car. The sky above is a clear blue, dotted with fluffy white clouds. The overall style of the video is dynamic and fast-paced, capturing the thrill and excitement of a race."
+    ],
+    "video_ids": [
+        "UwkuENf1GYU_11_0to195"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red car driving on a racetrack, which matches the 'Object(s): Red car' condition. The car is visible throughout the sequence, and there are no elements contradicting this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a red car driving on a racetrack. which matches the descriptionObject(s)' Red car' condition. The car is the in the frames, and there are no additional ining this description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..936087846daa7eaa77e79714449cdc079d0d3e56
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c4e3a77855052e0c7e312fd9e380a4ca0f0b2c33b4016860af25d0c0152ce40f
+size 269335
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..0c779052563cad1ed07ca9f4626f357e5c27b1db
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a pile of colorful clothing items, including shirts, pants, and sweaters, in various patterns and colors. The pile appears to be in motion, with the clothes shifting and moving around each other. The style of the video is dynamic and vibrant, with a focus on the textures and colors of the clothing. The camera angle is low, looking up at the pile of clothes, which adds to the sense of movement and energy in the scene. The video captures the essence of a busy, colorful, and chaotic environment, possibly a thrift store or a clothing donation center."
+    ],
+    "video_ids": [
+        "MFpNYQd2t64_11_0to120"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Shirts, pants, and sweaters.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a pile of clothes that includes various garments such as shirts (e.g., the blue satin-like shirt), pants (e.g., the denim jeans visible at the bottom), and sweaters (e.g., the beige knitted sweater on the left). These items match the 'Object(s)' condition of shirts, pants, and sweaters.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts a collection of colorful with appear various garments such as shirts,in.g., green green,-like fabric), pants (e.g., the green-like), in the bottom), and sweaters (e.g., the red sweateritted sweater). the right). The items are the 'Object(s)' condition of shirts, pants, and sweaters,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8c2d0fee9febd0dfc8283ae5c1ab1deebd1ba541
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:adcb252fd9e7f6708a13561a1052c030d0fd67d3b9d8e76ab415132c2df2cc1e
+size 127969
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..518f9987d2dc6631142359dd524ecd0362b3a0ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a blue shirt standing in front of a brick wall. He has a beard and mustache, and his hair is brown. The man is speaking, and his expression is serious. In the background, there is a painting of a cow on the wall. The style of the video is straightforward and appears to be a casual interview or conversation. The focus is on the man and his speech, with the background serving as a simple, unobtrusive setting."
+    ],
+    "video_ids": [
+        "0FZzoydAD8M_154_77to217"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a brick wall, a painting of a cow? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a blue shirt, a brick wall in the background, and a painting of a cow (partially visible on the upper left). These elements match the 'Object(s)' condition described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features shows a man in a blue shirt, a brick wall in the background, and a painting of a cow.partially visible). the right right side The elements match the descriptionObject(s)' condition provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..df3b4c9bb19bbba8715f53453cf1c3336bdbfa29
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f75ba25f487b866021977614f3d08c47d9f3cf7ee5976b14aa92c7358a6c905a
+size 242146
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..785ebb97ff54dac2b959bcf87ae33d4332630974
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a small, narrow river flowing through a rocky landscape. The river, with its calm and clear waters, is surrounded by large rocks and boulders on both sides, creating a natural corridor for the water to flow through. The rocks are of various sizes and shapes, adding to the rugged beauty of the scene. The river is flanked by trees with red leaves, adding a splash of color to the otherwise monochromatic landscape. The leaves of the trees are scattered along the river, creating a beautiful contrast against the green of the trees and the blue of the water. The video is shot from a high angle, providing a bird's eye view of the river and its surroundings. The overall style of the video is naturalistic, capturing the beauty of the landscape in its raw and unadulterated form."
+    ],
+    "video_ids": [
+        "OJlgfbDSVxA_51_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: River, rocks, trees with red leaves, scattered tree leaves, river water? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a river flowing between rocks, with trees displaying red leaves, scattered tree leaves on the ground and water, and visible river water with reflections. All specified elements are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a river, through rocks, with trees that red leaves. scattered tree leaves floating the ground, in, and the river water. reflections. The elements elements are present and contribute with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ad86bee92357915acf214da6c4f8bc9b679ed3c7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:562709c3fae18a9287b84b6e8f7c78c43be465495d4457042b3a34525b609a7d
+size 376916
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..6cdeac4cc85c646cc0d9468c05987200654c6257
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The image is a promotional poster for the Marvel movie \"Avengers: Age of Ultron\". It features a group of superheroes standing in a line, each with their own unique costume and weapon. The characters are arranged in a way that suggests they are ready for battle. The background is dark and ominous, with a hint of a stormy sky, which adds to the dramatic effect of the poster. The characters are all looking towards the viewer, giving the impression that they are ready to face whatever challenges come their way. The overall style of the poster is dynamic and action-packed, reflecting the high-stakes adventure that the movie promises."
+    ],
+    "video_ids": [
+        "gxv2siKPP3A_2_20to218"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A group of superheroes with unique costumes and weapons.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video displays a group of superheroes, each wearing unique costumes and holding distinct weapons (e.g., Thor with his hammer, Iron Man with his gauntlets, Captain America with his shield, Black Widow with her weapon, Hulk without a weapon but clearly identifiable, etc.). The characters are clearly identifiable as superheroes from popular media, and their attire and equipment are unique to each individual, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a group of characters with each wearing unique costumes. standing weapons weapons oror.g., a's a hammer, Captain Man with his repuntletlets, and America with his shield). and Widow with her bow, and with a weapon but in a by and.). The characters are standing designed as superheroes, their franchises, and their costumes and equipment are distinct to their character, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6911eee693c8167c2325ffb89c9d6fdb9dc238d9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb7104c4fdd21ca1359c115beb4e6a753d987ef08ce0102904ecc9721690ce23
+size 138060
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..97665ffd82de9f84b36065d8d6b260672a3d1465
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is an aerial view of a highway and surrounding area. The highway is busy with various vehicles, including cars, trucks, and motorcycles. The surrounding area includes commercial buildings, parking lots, and green spaces. The highway is elevated, with a retaining wall on one side. The sky is clear and blue, suggesting a sunny day. The style of the video is realistic, with no visible filters or artistic effects. The focus is on the movement of traffic and the layout of the urban environment."
+    ],
+    "video_ids": [
+        "nN3vYorHLwQ_3_0to104"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Various vehicles (cars, trucks, motorcycles), commercial buildings, parking lots, green spaces? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows various vehicles including cars and trucks moving on a multi-lane highway. Commercial buildings with flat roofs are visible adjacent to the highway, along with parking lots filled with parked vehicles. Green spaces with vegetation are also present along the highway and in the background, fulfilling all elements of the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts various vehicles, cars and trucks on on a multi-lane road. There buildings are distinct roofs and visible on to the highway. and with green lots and with vehicles cars. There spaces, trees are also present on the sides. in the surrounding. fulfilling the the of the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5bcfd6fed11c2f0887ba19d1995d728598434b3f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1244-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45cb8e72419b3fd215b1e9e1c421c54b04598bab447d5ce1b882b1fa88e969ee
+size 192506
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..009d3888ba9bc2345337810834646357c36a4154
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment on a football field, featuring a man in a black and orange jacket with the word \"Cincinnati\" and \"Bengals\" written on it, and a man in a black and orange uniform with the number 99. The man in the jacket is wearing a headset and appears to be speaking to the man in the uniform. The background shows a blurred crowd of spectators, indicating that the scene is taking place during a football game. The style of the video is a standard sports action shot, capturing the interaction between the two men on the field."
+    ],
+    "video_ids": [
+        "-mJl6HN1YLk_24_105to248"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black and orange jacket and a man in a black and orange uniform with the number 99.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a black and orange jacket (a coach) and another man in a black and orange uniform with the number 99 (a player). Both individuals are visible and match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two man in a black and orange jacket andlikely coach) and a man in a black and orange uniform with the number 99 (a player). The are are wearing and match the description provided.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6e4ec73f036c0f05428e8bb619916d31888d5ff0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:20b48d9b9b9d3f60e4415bb81a83643d096e2471a5ef0108af4c51ad76478483
+size 183432
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..e6e774eeeeb7cbd8e49f7f3edf8c50a92309474f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are standing in a well-lit, spacious closet filled with various clothes and accessories. The man, with curly hair, is wearing a white shirt and a silver chain necklace. The woman, with blonde hair, is holding a phone and appears to be showing something to the man. The closet is organized with white shelves and drawers, and there are several handbags and shoes visible. The overall style of the video is casual and intimate, capturing a moment of interaction between the two individuals in a personal space."
+    ],
+    "video_ids": [
+        "26x_uWD8uz0_27_0to205"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman interacting in a closet setting. The man is in the foreground, speaking to the camera, while the woman is in the background, holding a phone and engaging with him. Both individuals are distinctly visible and central to the scene, fulfilling the 'Object(s): A man and a woman' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a man and a woman standing with what room-like. The man is holding the center, and, the woman, while the woman is partially the background, holding a phone. a with the. The individuals are present human and fit to the scene, fulfilling the 'Object(s)' A man and a woman' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f934581f1ee05b42edf01fe6cd05a3616cf42783
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c5c539eac8e2d834fb5a69c07ae78b2e0421c44a35bdfd2bf554c7c070faada
+size 165006
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..ab00ceeff3d868290b05a90d16995964eb5cfda9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young woman with long brown hair is seen in a brightly lit room with a green wall. She is wearing a white and yellow top and has a black choker around her neck. She is smiling and appears to be in a cheerful mood. In the background, there are two children, a boy and a girl, who are also smiling and seem to be enjoying themselves. The boy is wearing a red shirt and the girl is wearing a green shirt. The room has a playful and colorful atmosphere, with various objects scattered around, including a vase, a bottle, and a sign that reads \"Your life is now. Seize it and make it amazing.\" The video seems to capture a happy and positive moment between the woman and the children."
+    ],
+    "video_ids": [
+        "6HRmjLM3Gf4_1_367to540"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Young woman, two children (a boy and a girl), vase, bottle, sign? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a young woman with two children (a boy and a girl) positioned closely together. In the background, a vase and a bottle are visible on a shelf, along with a sign that reads 'YOUR LIFE IS NOW SEIZE IT AND MAKE IT AMAZING'. All specified objects are present and accurately depicted, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows depicts a young woman and long children,a boy and a girl). in in together. The the background, there vase and a bottle are visible on a shelf, and with a sign. appears 'Sil'.'. YOUR'.ASON THE.' MAKE IT GREATING.' The elements elements are present and do depicted in fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..06aae6f618d9f61599fbbe11fa61fa85a2f41e20
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37e363d2358f8102f47d9ee6376361402aa8afecd971ef1158d1299bd062c1e7
+size 158323
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..7f6e0013f244a2c7ecbab089a460e214991391f4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a car review featuring a man in a suit and tie sitting in the driver's seat of a luxury car. The man is wearing glasses and appears to be speaking to the camera, likely discussing the car's features and performance. The car's interior is visible, showcasing the dashboard, steering wheel, and various controls. The car is parked in a showroom or exhibition area, with other cars visible in the background. The lighting is bright and even, highlighting the car's design and the man's professional attire. The style of the video is informative and polished, aimed at providing viewers with a detailed look at the car's interior and the man's expertise."
+    ],
+    "video_ids": [
+        "gYay3bPFMGw_14_0to133"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a suit and tie, glasses, luxury car (interior including dashboard, steering wheel, and controls).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a suit and tie, glasses, sitting inside a luxury car. The interior of the car, including the dashboard, steering wheel, and controls, is clearly visible. The scene is consistent with the described 'Object(s)' condition without any contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit and tie, glasses, and inside a luxury car. The interior of the car is including the dashboard, steering wheel, and controls, is visible visible. The man does consistent with the description 'Object(s)' condition.\"\n any contradictions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..276b8930540bd844fdce51ae3df3ea474bd6045b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c9af525460dd9494cd8c92ff6d28b1501d96fe35fc8ebabadaf7dfa43045076
+size 194918
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..c388c21b83faea882035951c894dc7f91b4b0a9b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a professional, corporate-style video featuring two men in suits. The first man is standing with his hands in his pockets, looking at the second man who is standing with his hands on his hips, looking at the first man. The men are in a modern office setting with white walls and large windows. The video has a blue overlay with white text that reads \"9. YOUR FRIENDS WANT THE SIMPLE LIFE.\" The overall style of the video is sleek and professional, with a focus on the interaction between the two men."
+    ],
+    "video_ids": [
+        "KiOgWjClYGs_44_65to272"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men in suits.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men dressed in suits walking and talking in an office setting, which matches the 'Object(s): Two men in suits' condition. The additional elements, such as the text overlay and background, do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two men dressed in suits, side interacting in a office setting. which align the descriptionObject(s)' Two men in suits.' condition. The presence elements, such as the office on and the, do not conflict the description description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3056c6ea5cb42a981d055604e22a1dff6e105897
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:55d99f6268796df7b9cf002915f8d8651d816ad4e93965d7a38bdece67162d5b
+size 100987
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..58eedc14bb3bec29bdf57989a3cf0b6b429b467b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases an intricate arrangement of colorful dominoes on a wooden floor, forming a large, detailed scene. The central focus is a black square surrounded by white dominoes spelling out \"THE DUCK\" in bold letters. Surrounding this central piece are various patterns and shapes created with blue, yellow, red, and clear dominoes, forming a vibrant and dynamic composition. The camera pans across the scene, revealing the complexity and creativity involved in the arrangement. As the camera moves, it captures the meticulous details of each domino piece, highlighting the artistic effort behind the design. The background features a row of white dominoes standing upright, adding to the overall structure and symmetry of the scene. The video concludes with a close-up shot of the \"THE DUCK\" text, emphasizing the central message amidst the colorful chaos."
+    ],
+    "video_ids": [
+        "f6cdde930e697c819eab71747dd5d28673308b6fb6d4ed44a85b9fc6bd18a071"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dominoes (black square, white, blue, yellow, red, clear), wooden floor, upright white dominoes.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a large arrangement of dominoes on a wooden floor, including black, white, blue, yellow, red, and clear dominoes. There are also upright white dominoes visible in the background. The setup matches the described elements without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows shows a variety number of colorfulinoes in a wooden floor. with black, white, blue, yellow, red, and clear piecesinoes. The are also upright white dominoes visible in the background. The presence matches the description ' without any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8250243bf376e2cc41be902165651c7704b33c2b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4842a75335fea8eab8a80d49a6ba72f003e28bdffa2efe691ecc8e7a2e71eb22
+size 141109
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..35a81a6c77b0b614dfa3316e9d05c1b48601fa0b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a collection of chocolate bone-shaped dog treats on a white plate. The treats are arranged in a pile, with some overlapping each other. The treats have a rich, dark brown color and are designed to resemble bones, complete with holes for added realism. The plate is placed on a wooden table, which provides a warm, natural contrast to the cool, smooth surface of the treats. The overall style of the video is simple and straightforward, focusing on the treats themselves without any additional embellishments or distractions. The lighting is soft and even, highlighting the texture and color of the treats without creating harsh shadows or glare. The video is likely intended for a pet-related context, such as a pet store or a pet owner's social media post."
+    ],
+    "video_ids": [
+        "peUy20IyzzQ_14_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chocolate bone-shaped dog treats arranged in a pile.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a pile of chocolate-colored, bone-shaped dog treats arranged on a white plate, which matches the description. The treats are clearly identifiable as dog treats shaped like bones, and they are piled together on a plate, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a plate of chocolate bone objects bone-shaped objects treats arranged on a white plate. which align the description of The treats are consistent visible as dog treats due like bones, and the are indeed together, a plate, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..142ea5556c29d35ab54278d0295eed8984842157
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2225ac1c8b2f05b13d15505a180e3f42f2939992a11db7da96c1ac5680840b0
+size 38528
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..324e607ce5c08a25dca20faf6109ba12a4d92ff6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in front of a promotional poster for a boxing event. The man is wearing a black t-shirt and a black baseball cap with white lettering. He has a beard and mustache, and he is smiling at the camera. The poster behind him features two boxers and a woman, all of whom are depicted in a fighting stance. The boxers are wearing boxing gloves and shorts, and the woman is wearing a black dress. The poster is colorful and features the names of the boxers and the event. The man appears to be a fan or supporter of the event, as he is posing for the camera with a smile. The style of the video is casual and informal, with the man appearing relaxed and comfortable in front of the camera. The focus of the video is on the man and his interaction with the camera, rather than the poster or the event itself."
+    ],
+    "video_ids": [
+        "MZnpmHkyT9w_13_0to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a promotional poster featuring two boxers and a woman, the boxing event details on the poster.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man in the foreground, and behind him is a promotional poster that includes images of two boxers and a woman. The poster also displays text that appears to be event-related, matching the description. There are no elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man standing the foreground and which in him, a promotional poster that includes two of two boxers and a woman. The poster also contains text that appears to be related details details such the ' of The are no additional in contradict the given description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b1022632a6032e3bc33128832e7c7ccaeb2be789
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1245-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a5e100da60974635c4dfba05d60a0aa89b7823ca36a3e06d1ced41d2d3a8b8d
+size 168729
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..b281be43a64e04c27425dd5629779dbbf99475d1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a black hoodie and a baseball cap sitting at a wooden dining table. He is eating a meal, which includes a plate of food and a glass of water. The table is set with a red placemat, a fork, a knife, and a spoon. There is also a candle in a glass holder on the table. The man is sitting on a wicker chair. The setting appears to be a home dining room. The style of the video is casual and everyday, capturing a simple moment of a person enjoying a meal."
+    ],
+    "video_ids": [
+        "cCJ8dYm3LQk_16_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black hoodie and baseball cap, a wooden dining table, a plate of food, a glass of water, a red placemat, a fork, a knife, a spoon, a wicker chair, and a candle in a glass holder.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black hoodie and baseball cap seated at a wooden dining table. There is a plate of food in front of him, along with a fork, knife, and spoon. A red placemat is visible under the plate, and a candle in a glass holder is on the table. A wicker chair is also present. While a glass of water is not clearly visible, the core elements described are largely present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person in a black hoodie and a cap, at a wooden dining table. On is a plate of food in front of him, a with a glass, knife, and spoon. A glass placemat is under under the plate, and a glass in a glass holder is present the table. A glassicker chair is also present in The there glass of water is not visible visible, the presence elements of in present present and match with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ee077292301d3b6c23f0759f51162f19dbac113b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7353d44e52c7fd54e709b2eea50c567b86459235402ed645138dca06e04a1586
+size 210763
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..6cf1b34e54024c8cc05775b90970fa58b9c08cde
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a white goat's head, which is positioned against a wooden fence. The goat has two prominent horns, one of which is slightly curved. Its ears are large and pointed, and it appears to be looking around its environment. The goat's fur is predominantly white, with some brown patches visible on its face and horns. A chain is attached to the wooden fence, suggesting that the goat is tethered or confined in some way. The background includes a glimpse of what seems to be a barn or stable, with hay scattered on the ground. The lighting in the video is natural, indicating that the scene takes place during the day. The goat remains relatively still throughout the video, occasionally moving its head slightly, possibly reacting to sounds or stimuli in its surroundings. There are no significant changes in the goat's position or the environment, maintaining a consistent focus on the animal and its immediate setting."
+    ],
+    "video_ids": [
+        "401262f423b62e6e50cb05fb95c133c2142e7d980bd845ca4c243025f5c11534"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: White goat's head, wooden fence, chain, barn/stable, hay? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a white goat's head with visible horns, positioned near a wooden fence with a chain attached. The background suggests a barn or stable setting, and hay is visible on the ground, matching the described elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a white goat's head, prominent horns, a in a wooden fence with a chain.. The background suggests a barn or stable setting, and there is visible at the ground, indicating the description elements.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e3cb09e8b2c72fb13d62fdc87cf22dfb8b8bfa49
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:42e8040074a3e82bb9034a223243bd55477e910c818d8287538303add9368d5a
+size 221927
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..7ac756465ffa143bee56b7b602abf332cc5f9091
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a close-up view of an Audi engine, focusing on the TFSi logo and the red accents. The style of the video is a high-resolution, professional photograph that captures the intricate details of the engine. The engine is black with red accents, and the TFSi logo is prominently displayed. The video is likely intended for promotional or informational purposes, highlighting the features and design of the Audi engine."
+    ],
+    "video_ids": [
+        "X0OAuAdtqHM_53_30to201"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Audi engine, TFSi logo, red accents? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows an Audi engine with the TFSI logo prominently displayed on the engine cover. The engine cover also features red accents, matching the description. All core elements mentioned in the condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows an Audi engine with a TFSi logo prominently displayed. the cover cover. The presence cover also features red accents, which the description. The elements elements of in the description are present in accurately depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cdd29aa9220496143904b0223ff4c04d94d0b381
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ad6d1a3085d1d40974fa06e75ee5caca794b24b9da255a9207aa270ed497bfb5
+size 207544
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..76aee70a55bac434a9a2623dea106f8fda4b9558
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are engaged in a conversation. The man, dressed in a suit and tie, is seated comfortably in a chair, his hand resting on his chin as he listens attentively to the woman. The woman, seated opposite him, is actively engaged in the discussion, her body language suggesting a serious or intense conversation. The setting appears to be a cozy living room, with a bookshelf filled with various books in the background, adding a touch of intellectual charm to the scene. The overall atmosphere of the video suggests a thoughtful and meaningful exchange between the two individuals."
+    ],
+    "video_ids": [
+        "_KR1wPK8T4Q_8_0to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two individuals: a man in a suit and tie, and a woman with gray hair seen from behind. Both are present and engaged in what appears to be an interview or conversation, fulfilling the 'A man and a woman' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a individuals, a man in a suit and a and and a woman in her hair, from the. The are seated in engaged in a appears to be a interaction or discussion setting fulfilling the 'Object man and a woman' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ebbac700aa1c80f735d029a74cb78f221743a000
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b8c3778e15a639e74eed884ec7b2f50086b0304e116f33f51f31c12d3646cfa1
+size 124942
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..9c3b13de3f9ae0ce80b61bf2e2284b562036ff4a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red toolbox with various automotive fluids and tools on top of it. The fluids include two bottles of blue Peak antifreeze, a white bottle of water, and a red bottle of oil. A yellow funnel is also present on the toolbox. In the background, there are three cars with their hoods open, indicating that they are being worked on. The style of the video is a straightforward, real-life depiction of a car repair or maintenance scenario. The focus is on the toolbox and the fluids, with the cars in the background serving as context for the activity. The video does not contain any stylized elements or artistic effects."
+    ],
+    "video_ids": [
+        "29z94xBdAjE_13_853to980"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red toolbox, two blue Peak antifreeze bottles, one white water bottle, one red oil bottle, one yellow funnel, three cars with open hoods? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a red toolbox with various items on top, including two blue Peak antifreeze bottles, one white water bottle, one yellow funnel, and a red oil bottle (though the oil bottle is not clearly visible, the red cap and label suggest it). In the background, three cars with open hoods are visible, matching the description. All core elements are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows ful a red toolbox with a items on top, including two blue Peak antifreeze bottles, one white water bottle, one red funnel, and three red oil bottle.which the red bottle is not clearly visible). it red color suggests shape suggest it is Additionally the background, there cars with open hoods are visible, fulfilling the description. The elements elements are present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..908097fae513fd0444566c7c31412ae6ec42b8f6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:36ce4a733eb93504f8019629c318b119c06fced566ef0aaf14084cf1af9dffa9
+size 80513
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..a9bdaf152518ab06f4027038860a79799f011cd6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and a baseball cap, sitting in a dark room with a brick wall in the background. The man is wearing a gray shirt with a pattern of small white dots. The lighting in the room is dim, with a soft glow coming from the left side of the frame. The man appears to be in deep thought or contemplation, as he gazes off to the side. The overall style of the video is moody and introspective, with a focus on the man's facial expression and the texture of the brick wall in the background."
+    ],
+    "video_ids": [
+        "EjQ8BEr4cYs_0_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and baseball cap, wearing a gray shirt with small white dots.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and a baseball cap, wearing a gray shirt with small white dots, which matches the description. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and a baseball cap, wearing a gray shirt with small white dots. which matches the description provided The presence appears additional elements in not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ab9ccd7f2b75609eb9305ff7b12d3e4f047ae79c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f36f21d7494fd36b912c8244f4c25652355168751168b5a20bd5b0fba7cf7595
+size 156065
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..a8b6b6ef207b2d8374369eb7a21f8b457a00cd51
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a yellow sports car parked in a parking lot. The car is customized with a black stripe and a \"Fake Taxi\" logo on the side. The car is equipped with large silver rims and a black front bumper. The car is parked next to a black car and a blue car. The parking lot is filled with other cars, indicating that it is a busy area. The car is the main focus of the video, with its vibrant yellow color and unique customization standing out against the other cars in the lot."
+    ],
+    "video_ids": [
+        "K6MjYdIc6TA_5_0to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Yellow sports car, black car, blue car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a yellow car, which matches the 'Yellow sports car' description. In the background, there are also black and blue cars visible, fulfilling the 'black car, blue car' condition. The video does not contradict these core elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a yellow sports, which is the descriptionYellow sports car' part. Additionally the background, there are black black and blue cars,, which the 'black car' blue car' condition. The presence does not contradict the elements elements,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..58ec91193df92f2f1e5570cd3963633bc03a8615
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48132f344dfc13b435fd9b69e4b1fbcf9060013c5aa67e861375ed49f031f0b6
+size 183273
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..789d4244b1829f95023676544e0db6b0e15ee766
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a bowl of colorful cereal, specifically Froot Loops, in a blue bowl. The cereal is vibrant and appears to be freshly poured. The bowl is placed on a surface, possibly a table or countertop. The focus is on the cereal, with no other objects or actions taking place. The style of the video is simple and straightforward, with no additional elements or distractions. The video is likely meant to showcase the cereal, possibly for advertising or promotional purposes."
+    ],
+    "video_ids": [
+        "GJ73UdG6r40_41_23to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bowl of colorful Froot Loops cereal.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a bowl filled with colorful Froot Loops cereal, which matches the core description. The cereal pieces are distinctly colorful (pink, yellow, blue, white) and ring-shaped, consistent with Froot Loops. The bowl is turquoise, and the cereal is being poured into it, fulfilling the 'bowl of colorful Froot Loops cereal' condition. Additional elements like the person pouring the cereal and the background do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows ful a bowl filled with colorful,root Loops cereal. which matches the description description. The vibrant pieces are vibrant shaped andpink, green, green, green, and shaped-shaped, resembling with theroot Loops. The presence is blue, and the focus is piled interact, it, which the 'Objectowl'' Froot Loops cereal' condition. There elements like the hand's the cereal do the background are not contradict the main.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..de1055badbce6d01bb13e213546caccab3a7674f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1246-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8495f8f7de134ac06c3e12a426d0ca47d43526024640d9dc201bb35fe5e12535
+size 153805
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..1d086d1959d84a70c56548ff6e710fe7406e2a6e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two scientists in white lab coats and blue goggles are engaged in a discussion. They are standing in a lab with a greenish hue, and there are shelves in the background. The first scientist is holding a tablet, which he is showing to the second scientist. The second scientist is looking at the tablet with interest. The scientists are standing close to each other, indicating a collaborative and focused discussion. The overall style of the video is realistic and professional, capturing the essence of scientific research and collaboration."
+    ],
+    "video_ids": [
+        "xYRxfh76hhA_29_0to168"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two scientists in white lab coats and blue goggles? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men dressed in white lab coats and wearing blue goggles, consistent with the 'Object(s)' condition. Additional elements such as background equipment and a third person are present but do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two individuals wearing in white lab coats and wearing blue goggles, which with the descriptionObject(s)' condition. They elements such as the colors and the tablet person in not but do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..99228bc18720aa6d92ada790d1384bdea813735e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b097118d80627be41881b4c9605d6a881ed1c80b993631e80d4f8a28ab90d2fe
+size 121278
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..7b294825b9c735532f9461c337d2230d0501b7f7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit standing outside a building. He is wearing glasses and a tie, and appears to be speaking to a microphone. The man is identified as Michael Gove MP, Environment Secretary. The style of the video is a news report, with a focus on the man and his surroundings. The building in the background has a sign that reads \"Seacole Buildin\". The man is the main subject of the video, and his actions and words are the primary focus. The setting is an urban environment, and the man is dressed in formal attire, suggesting a professional context. The video likely contains information about the man's role as Environment Secretary and his views on environmental issues."
+    ],
+    "video_ids": [
+        "PLDmUYR50oE_15_0to181"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit, wearing glasses and a tie, speaking into a microphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed in a suit, wearing glasses and a tie, speaking into a microphone. The setting and his attire align with the description, and there are no elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in in a suit, wearing glasses, a tie, standing into a microphone. The setting appears attire attire match with the description provided and there are no additional in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..56b16cad116bb3a6f437c08c69fdad2032dcfc0f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:359d0a2c9db6eda8b507729b2a4570197e971cfdd545bcf06584af966979b18b
+size 143439
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..2e5fc7f1650bd0ccaf170951d81ddba98ad727da
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up shot of a meal being prepared, focusing on the ingredients and the process of cooking. The meal consists of a variety of vegetables, including green beans, lima beans, and lemon wedges, as well as chicken and rice. The ingredients are arranged in a bowl, and the camera captures the process of mixing and cooking the ingredients. The style of the video is realistic and informative, providing a detailed look at the preparation of a healthy and nutritious meal. The video is likely intended for viewers interested in cooking or looking for new recipes to try."
+    ],
+    "video_ids": [
+        "NVgEf_Zpkes_8_72to199"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Ingredients (green beans, lima beans, lemon wedges, chicken, rice), bowl, cooking utensils (not explicitly mentioned but implied).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bowl containing ingredients such as green edamame (often mistaken for green beans or lima beans), a lemon wedge, shredded chicken, and what appears to be quinoa or rice. There are no cooking utensils visible, but their absence does not contradict the condition since they are not explicitly required to be shown. The core ingredients and bowl are clearly present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bowl containing green such as green beansamame,which mistaken for green beans), lima beans in lemon lemon wedge, and chicken, and rice appears to be riceinoa or a. The is no cooking utensils visible, but their implied does not contradict the description as the are not explicitly mentioned. be shown. The presence elements and the are present depicted,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..86ee8e5156ef301bdd45b772a4eafe5dc4651f1d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e6d334725fdc46a63f0a5dd2fbdab7df6c8597d74f4538b9d857f06dbaf64f03
+size 93044
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..86ec0d8fd418a5352e8a8aaf078d76f433353381
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a vibrant orange Mini Cooper car with a black stripe on the hood, parked on a street. The car is facing the camera, and its license plate reads \"DFD-45 07\". The car is parked in front of a lush green palm tree, which adds a tropical vibe to the scene. The car's shiny exterior reflects the sunlight, highlighting its sleek design. The car's black and orange color scheme stands out against the natural backdrop. The overall style of the video is bright and cheerful, with a focus on the car and its surroundings."
+    ],
+    "video_ids": [
+        "K_KcfLVHx0w_11_40to177"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A vibrant orange Mini Cooper car with a black stripe on the hood, license plate 'DFD-45 07'.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a vibrant orange Mini Cooper with a black stripe running down the hood, and the license plate is visibly 'DFD-45 07'. The car is the central focus and matches the description exactly.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows depicts a vibrant orange Mini Cooper car a black stripe on across the hood. which the license plate reads visible 'DFD-45 07'. The car is parked central focus, matches the description provided.}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..90074052310cf57d3dcedc8418e472d385bae0a0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0b8763da041c95d9cfa5bc858907c56b553b2bf3761c5407514ee79c7f03274
+size 206559
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..348d78010026084e886f495cc5650388e5ee03a2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a loaf of bread being sliced on a cutting board. The bread is a golden brown color, indicating it is freshly baked. The cutting board is placed on a wooden table, which provides a rustic backdrop for the scene. The bread is sliced into even pieces, and the slices are stacked neatly on top of each other. The video captures the process of slicing the bread, from the initial cut to the final stack of slices. The style of the video is simple and straightforward, focusing on the bread and the cutting process without any additional elements or distractions."
+    ],
+    "video_ids": [
+        "WVaVM9hJcX4_2_0to150"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A loaf of golden brown bread and a cutting board.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a loaf of golden brown bread with slices cut from it, placed on a white cutting board. The cutting board is on a wooden surface, and the background is green grass, which does not contradict the core description. All key elements of the object condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a loaf of golden brown bread being a being from it, placed on a cutting cutting board. The bread board is wooden a wooden surface, which the bread is blurred,, which is not contradict the core description. The elements elements of the description condition are present.\"\n match depicted.\"\n}<|im_end|>\n addCriterion",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fb20c0a1eb3c74f5d13b3fc0f92c537f93e39368
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3ed1574bf13c6336f3c6e993343db54fce04ec66226a96b0943b8998daaa05dd
+size 157788
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..cbbb9e35460ce16d46a247c8b949c6e84b1ef811
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two women engaged in a conversation. The woman on the left is wearing a colorful, patterned shirt and has blonde hair. She appears to be listening intently to the woman on the right, who is wearing an orange shirt and has dark hair. The setting is outdoors, with a building and grass visible in the background. The style of the video is casual and candid, capturing a moment of interaction between the two women."
+    ],
+    "video_ids": [
+        "AtYDWwpBiPA_8_0to174"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two women, one in a colorful, patterned shirt with blonde hair, the other in an orange shirt with dark hair.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two women standing outdoors. The woman on the left has blonde hair and is wearing a colorful, patterned shirt, while the woman on the right has dark hair and is wearing an orange shirt. These details match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two women, side. One woman on the left is blonde hair and is wearing a colorful, patterned shirt, while the woman on the right has dark hair and is wearing an orange shirt. The details match the description provided, the questionObject(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0333fb318a53e15c96231526a914d996a46b6041
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1ed25e3b044a49426ee29b60bb00536f15fbf9946cfa0b1b1f1fcd63a27bcff
+size 166755
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..545ad932eeb75fb03b0a0f00b729765b973b68c7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, three friends are seen enjoying a coffee break on a city street. They are all holding cups of coffee, with the cups featuring a distinctive red and white design. The friends are dressed in winter clothing, suggesting that the weather is cold. The street they are standing on is lined with buildings, and there are cars parked along the side of the road. The friends are smiling and appear to be having a good time. The overall style of the video is casual and friendly, capturing a moment of camaraderie among the friends."
+    ],
+    "video_ids": [
+        "_YGZkaJTG7U_84_0to153"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three friends, three cups of coffee with a distinctive red and white design, cars parked on the street.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three friends standing together, each holding a cup of coffee with a distinctive red and white design. Cars are visible parked on the street in the background, matching the description. Additional elements like a 7-Eleven store and a newspaper stand do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows two individuals, on, each holding a cup of coffee with a distinctive red and white design. There are parked parked on the street in the background, which the description. The elements like the building4-Eleven logo sign a building stand are not contradict the core description and}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..eb2580ea9612b3b435126a4441c4f7f361502e68
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c1635e914de6ea7dfd2cf1d77c90c46203f61d9ba673e0274a82168beb454f65
+size 269458
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..af7e2f0fa80bb6f2a7abb6334321a9309fd79b58
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a brown-skinned doll with long brown hair, wearing a pink and black floral dress. The doll is positioned in a room with a blue dresser and a television screen displaying a cartoon. The doll appears to be looking up, possibly at the ceiling or a light fixture. The room has a playful and colorful atmosphere, with the doll as the central focus. The style of the video is a simple, straightforward depiction of a doll in a room, with no additional action or movement."
+    ],
+    "video_ids": [
+        "XhtcNiDboG0_33_0to148"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A brown-skinned doll with long brown hair, wearing a pink and black floral dress.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a doll with brown skin and long brown hair, wearing a dress with pink and black floral patterns, which matches the description. The background elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a doll with long hair and long brown hair, wearing a pink with a and black floral patterns. which matches the description provided The setting elements, not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ab9a66091a69e4ce3099b2708a8399d68752a195
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1247-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f76c25c38150768702d0e5e4d766125ee1749029b29523484dcdcd8537c5c33
+size 157150
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..2b83beab2a398f75770113e7016b71845524c46b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a professional basketball player in a purple jersey with the word \"Lakers\" on it, standing on a basketball court. The player is looking off to the side with a focused expression. The style of the video is a close-up shot of the player, capturing his facial expression and the details of his jersey. The background is blurred, emphasizing the player as the main subject. The lighting is bright, suggesting an indoor setting. The player's stance and expression suggest he is in the middle of a game or practice."
+    ],
+    "video_ids": [
+        "70DoiNdIrOw_5_0to152"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Professional basketball player in a purple jersey with 'Lakers' written on it.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a professional basketball player wearing a purple jersey with 'Lakers' clearly written on it, matching the core description. The player's attire, including the team name and color, is accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person basketball player wearing a purple jersey with 'Lakers' written written on it. which the description description. The player's attire and including the jersey name and jersey, align consistent depicted, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..63a340a9017a22fe466f7c307eba15ca6aeb517f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f01066a06b1efd887ae32f8e464ac2ac144544148bb513a505cebf98ba313f3b
+size 138799
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..02592b8bd8cdfcd221afcaf13f9e225c240de9d8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a suit sitting in a chair, engaged in a conversation with another man who is not visible in the frame. The man in the suit is wearing glasses and has a beard. The setting appears to be an indoor room with a plain wall in the background. The style of the video is a straightforward, candid conversation between two individuals, likely a professional or formal setting. The focus is on the man in the suit, with the other man's presence implied by the man's gaze and body language."
+    ],
+    "video_ids": [
+        "0wE1DwtNgDg_2_54to217"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit sitting in a chair, wearing glasses and having a beard? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a suit, glasses, and having a beard, sitting in a chair. These elements align with the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a suit, glasses, and a a beard, sitting in a chair. The elements match with the description provided, the questionObject(s)' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..32ff95d70326faf0847df2dd4b65c96f0443c4ba
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd48ca20255c215d91dc387bd6a04158e3230350be52ee78b31673646834888f
+size 171311
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..b9c9b60a8e5f3be3bf9ae0c9ffd4e1d7495ed000
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in an orange shirt and a red baseball cap, participating in a food challenge. He is seated at a wooden table with a plate of steak, a salad, and a roll. He is using a knife and fork to cut and eat the steak. The table also has a glass of water and a bottle of hot sauce. The man is wearing a necklace and has an earbud in his ear. The video has a timestamp at the bottom right corner, indicating that it is 02:43. The style of the video is casual and candid, capturing the man's attempt to complete the food challenge."
+    ],
+    "video_ids": [
+        "Vr7FpLYOf6s_11_0to161"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in an orange shirt and a red baseball cap, a knife, a fork, a plate of steak, a salad, a roll, a glass of water, a bottle of hot sauce, a necklace, an earbud.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing an orange shirt and a red baseball cap, using a knife and fork to eat a plate of steak. A salad, a roll, a glass of water, a bottle of hot sauce, a necklace, and an earbud are also visible on or near the table, matching the described objects. All elements are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing an orange shirt and a red baseball cap. which a knife and fork to eat a plate of steak with There salad is a roll, a glass of water, and bottle of hot sauce, and necklace, and an earbud are also visible on the near the table. matching the description objects. The elements are present and do with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2cad51397d301bfa421a9ea361154061cf62bf28
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e5a5a24f73125e32305f2789693e90b4bc62847075628396d556f3a1690e338
+size 139568
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..6e3ebf14ee99dfc7563388834cbb539b4a76d841
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a gray Audi car in a garage. The car is parked and stationary. The car has a license plate that reads \"LP ML 87\". The car has a sleek design with a shiny exterior. The garage has a white floor and a white wall. There is a sign in the background that reads \"Profis\". The car is the main focus of the video, and it is shown from a rear view. The video is a still image, and there is no movement or action. The style of the video is straightforward and simple, focusing on the car and its details."
+    ],
+    "video_ids": [
+        "iXCcS2_Kfp8_5_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A gray Audi car with a license plate reading 'LP ML 87'.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a gray Audi car from the rear, and the license plate is visible and reads 'LP ML 87'. The car's model and color match the description, and no elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows shows a gray Audi car with the rear view and the license plate reads visible and reads 'LP ML 87'. The car's design and color match the description, and there additional contradict the information description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3aee8c865a4551f9eb559b253b26cb2cc59b2bb4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1689b6d73886099e10368c704e86fed1282907fa0736214a53b8015bb7f0b4f6
+size 157651
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..7673999e205eec6a5105c9977a81a1edf2b9ee11
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video opens with a close-up shot of a monkey's face, its eyes wide and alert, as it peers through the branches of a tree. The camera then pans to the right, revealing a group of monkeys moving through a dense thicket of dry, leafless bushes and trees. The ground is covered with dry leaves and twigs, indicating a dry season or a savanna-like environment. The monkeys are seen running and jumping, their bodies agile and swift, navigating the terrain with ease. The camera follows them closely, capturing their movements from various angles as they move deeper into the underbrush. The scene is filled with the sounds of rustling leaves and occasional monkey calls, adding to the natural ambiance. The video concludes with the monkeys disappearing into the thick foliage, leaving the viewer with a sense of the wild and untamed nature of these primates in their natural habitat."
+    ],
+    "video_ids": [
+        "da69bf2b6019ffeef1722b03dcfdf7f21aef65dedf67335c364bf9dc2185c333"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A monkey and a group of monkeys? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple monkeys in a natural, wooded environment. Some are on the ground, while others are in trees, consistent with the description of 'A monkey and a group of monkeys'. The scene accurately depicts a group of monkeys, fulfilling the object condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a monkeys in a natural setting forest environment. The frames in the ground, and others are partially the, which with the presence of aA monkey and a group of monkeys.' The presence includes depicts a group of monkeys in fulfilling the ' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c47c77b7c11593938b4767a732ba5ac2dcdbb355
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04070f9b61ec63fdffa1bc96a5d75c7cc9925126f335f504db66fa034d7fc068
+size 280718
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..79ca367869ffad247cbe175c23eb669480a71def
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a stylish ensemble, including a black hat and a yellow jacket adorned with patches. He is seated in a chair, gesturing with his hands as he speaks. The man's attire and demeanor suggest a casual yet confident style. The background is a plain white, which puts the focus entirely on the man and his actions. The video appears to be a still image, as there is no indication of movement or change in the scene. The man's position in the chair and his active gesturing suggest that he is engaged in a conversation or presentation. The overall style of the video is simple and straightforward, with a focus on the man and his actions."
+    ],
+    "video_ids": [
+        "7wEEgUHJcrU_1_41to175"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a stylish ensemble (black hat, yellow jacket with patches), seated in a chair.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black hat and a yellow jacket adorned with patches, seated in a chair, which matches the core description. Additional elements like bracelets and the chair's design do not contradict the primary description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man seated a black hat and a yellow jacket with with patches, seated in a chair. which align the description description provided The elements such the and the background are presence do not contradict the main description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..07e2f8d78f5118bb4807019bd00a44599c0d3a62
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba4e612d6344a98df9a63fae0e4419728db4d9c7a623d187a6d79f795e9612fc
+size 160782
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..7e2a1ed221428c06b004c298fb0340101533b90b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a baseball player in action, pitching a ball on a field. The player is dressed in a blue and white uniform, with the number 3 prominently displayed. The player's stance and the motion of the ball suggest a powerful throw. The field is well-maintained, with a clear view of the pitcher's mound. The background is slightly blurred, emphasizing the player and the action. The video is likely taken from a professional game, given the quality of the footage and the player's uniform. The focus is on the player and the ball, with the rest of the field serving as a backdrop to the action. The video is dynamic and captures the intensity of the game."
+    ],
+    "video_ids": [
+        "qEayY7DmH30_2_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Baseball player in a blue and white uniform with the number 3, a ball being pitched.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young female baseball player wearing a blue and white uniform with the number 3, and she is in the motion of pitching a baseball. The core description is accurately represented, even though additional elements like spectators and a batter are present, which do not contradict the primary focus.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a baseball baseball baseball player in a blue and white uniform with the number 3, pitching she is in the process of pitching a baseball. The description elements of largely represented in and though the elements like the or the detailed are not, they are not contradict the main focus on}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..de09ab580dc0151f7db978c655a4115dac966741
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:02f63a5e27f6be582c590912e5fbf79256d51ab3dcf9d6ed223060269b839ab5
+size 166205
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..bb5a230b218f886d57f40d53b2a0a8addf19740b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man giving a presentation on a stage. He is holding a tablet in his left hand and gesturing with his right hand. He is wearing a blue t-shirt and has a microphone attached to his ear. The stage is lit with blue and white lights, and there is a large screen behind him displaying a slide. The man appears to be speaking to an audience, and his body language suggests that he is engaging with them. The style of the video is a straightforward presentation, with the man as the main focus. The lighting and stage design create a professional atmosphere."
+    ],
+    "video_ids": [
+        "23HjmMbgLLg_47_0to193"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a tablet, a microphone, a large screen displaying a slide.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man holding a tablet and wearing a microphone, standing in front of a large screen that displays a slide (with silhouettes). All core elements described are present, and no contradictions are evident.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a tablet, standing a microphone, which in front of a large screen displaying displays a slide.pres ahouettes and The the elements of in present, and there additional are observed.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f65a6e0484f264ee8198d38b3b34cbf2357df645
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1248-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d220d6d853f245439552c0bff0a565f381eecbf598b545022bad404bd051392c
+size 184191
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..06f70827885604a5bb563ebf6d1898ad9b07e75d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features an older woman with curly hair, wearing a blue shirt, sitting in a chair. She appears to be engaged in a conversation or interview. The setting is an office or a room with a bookshelf in the background. The woman is the main subject of the video, and her facial expressions and body language suggest she is actively participating in the discussion. The style of the video is a straightforward interview or conversation, with the focus on the woman's responses and the context of the setting."
+    ],
+    "video_ids": [
+        "VnEA39R3n74_38_63to197"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An older woman with curly hair, wearing a blue shirt, sitting in a chair.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an older woman with curly hair, wearing a blue shirt, and sitting in a chair, which matches the core description. Additional elements like a framed photo and bookshelf in the background do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an older woman with curly hair, wearing a blue shirt, sitting sitting in a chair. which matches the description description provided The elements such the book picture and ash in the background do not contradict the main and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..304f258bbf52d4166f8d4dada7c2f53bb7239ad0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79daa6d15c88afb0d0d6fc9028a71b57954961d8374def0e95cb6eaf56a7df6e
+size 136095
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..32c4819b7541c41902254a8714e8d389083152bf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with short brown hair, wearing a black and white striped dress, standing in a bustling city street. She is looking directly at the camera with a serious expression on her face. The background is filled with colorful lights and signs, indicating a lively urban environment. The woman is the main focus of the video, and her position in the frame suggests she is the subject of the video. The style of the video is realistic, capturing the everyday life of the city."
+    ],
+    "video_ids": [
+        "AGXyXmujyvw_12_78to213"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with short brown hair, wearing a black and white striped dress.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with short brown hair and a black and white striped dress, which matches the description. The background elements, such as blurred lights and pedestrians, do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with short brown hair wearing she black and white striped dress. which matches the description provided The background is, such as the lights and people, do not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f6a497176601650bec5729992f9be09ae67f70fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e26efebde6e41e50c5e83948a56f83fb8691066688554164549b49f9c0e002be
+size 205746
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..e74ff85d8617e260e3d3dc968d4a9b07d410456f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and short hair, wearing a plaid shirt. He is seated in front of a wooden wall with a large, red and white logo. The man appears to be engaged in a conversation or interview, as he is looking off to the side with a thoughtful expression. The lighting in the room is soft and warm, creating a relaxed atmosphere. The overall style of the video is casual and informal, with a focus on the man and his surroundings."
+    ],
+    "video_ids": [
+        "1wQ8FZQx54I_24_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and short hair, wearing a plaid shirt, engaged in a conversation or interview.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and short hair, wearing a plaid shirt, who appears to be engaged in a conversation or interview. His facial expressions and slight head movements suggest active participation in dialogue, and the background context (wooden wall with signage) supports an interview or casual discussion setting. There are no elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and short hair, wearing a plaid shirt, who appears to be engaged in a conversation or interview. The facial expressions and head head movements suggest interaction participation in a. which the background, supportsaen panel and a) is the interview or talk talk setting.\"\n The are no elements in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..efa2f455bc4d162ba093d93c3ab0da6c148b4040
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5fa707c13f58af7921609152258997355ae8f8e060cf9d6d6dfc541059b38c93
+size 210419
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..f56f09ec28226786eb8276f4a349e510235d79c4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video presents a sequence of three frames featuring a robot in a futuristic setting. The robot, with its sleek, metallic design, stands out against the dark, gray background. In the first frame, the robot is shown from a side angle, its body angled slightly forward as if it's in motion. The second frame offers a front view of the robot, showcasing its intricate details and design. The third frame provides a top-down perspective, emphasizing the robot's height and the complexity of its structure. The overall style of the video is sleek and modern, with a focus on the robot's design and movement."
+    ],
+    "video_ids": [
+        "fCUMv_7obC8_156_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A robot with a sleek, metallic design.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features multiple robotic figures with a sleek, metallic design, consistent with the description. The robots have polished, reflective surfaces and angular, futuristic forms, fitting the 'sleek, metallic design' criterion. Additional robots in the background do not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a frames figures with a sleek, metallic design. characterized with the description. The robots have a, reflective surfaces and angular, streamlined shapes, which the 'sleek, metallic'' criteria. The elements in the video do not detr the description but}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..07aab62e50421f8d516df3f0b0e36c710021cd52
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1229065bebc26780fc12753aea6f25a18072cba391b83aac1572e6230685a96f
+size 216823
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..c34a55e6cb44603683ca3827468f1ad77e50e3b6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a dynamic sports montage featuring a Minnesota Vikings quarterback in action. The quarterback, wearing a purple jersey with the number 3, is captured in three distinct moments: throwing a pass, celebrating a successful play, and preparing for the next play. The video is set in a stadium filled with enthusiastic fans, creating an atmosphere of excitement and anticipation. The quarterback's movements and expressions convey a sense of determination and focus, reflecting the intensity of the game. The video is likely edited to highlight the quarterback's skills and the team's performance, with a focus on action and emotion."
+    ],
+    "video_ids": [
+        "FIolhoXtyYA_53_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A Minnesota Vikings quarterback (wearing a purple jersey with the number 3).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a Minnesota Vikings quarterback wearing a purple jersey with the number 3, which matches the description. The player is in a football uniform, and the team name 'Vikings' is visible on the jersey. The presence of other players and the crowd does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a football Vikings quarterback wearing a purple jersey with the number 3. which matches the description. The player is depicted a football stadium, and the setting's andVikings' is visible on the jersey. The setting of a players in the football in not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ce75a6fecb1c43072069707c1bd4c1d07638aaf2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3bb01edcc282c4e9f8463684a588bfd1043b6366bf3f6efbb5738d6df2dd8f5
+size 260478
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..18d4261124f820758d9db0d6983ddc4f83697fe9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the process of grilling chicken and vegetables on a black cast iron skillet. The first frame shows the skillet being placed on the grill, with the chicken and vegetables just starting to cook. The second frame shows the chicken and vegetables cooking, with the chicken becoming golden brown and the vegetables becoming tender. The third frame shows the chicken and vegetables fully cooked, with the chicken being removed from the skillet and placed on a plate. The style of the video is a close-up, time-lapse shot that captures the transformation of the food from raw to cooked. The focus is on the food, with the grill and the surrounding environment being out of focus. The video does not contain any text or narration."
+    ],
+    "video_ids": [
+        "OXoPeCRNmek_4_585to725"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Black cast iron skillet, chicken, vegetables? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black cast iron skillet on a grill, containing chicken pieces and vegetables (onions and mushrooms), which matches the described objects. The additional elements, such as tongs and a plate, do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black cast iron skillet with a grill, with pieces pieces and vegetables.greenions and bell). which align the description objects. The presence elements like such as theongs and the blurred, do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d50cb9dfc25454530add6604def8007b1ca77911
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc648a6bda3cae9089844c237223b11ce7ce665dd18f741abee1a1caa0deb447
+size 239010
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..8dd8677575ddaff746c4617d6e2f02e5ea06f375
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is an aerial view of a lush green forested area, with a few scattered houses nestled among the trees. The houses are small and modest, blending in with the natural surroundings. The forest is dense and verdant, with a variety of trees and foliage. The sky is clear and blue, suggesting a sunny day. The overall style of the video is serene and peaceful, capturing the beauty of nature and the tranquility of the rural setting."
+    ],
+    "video_ids": [
+        "0f94w91lL2o_177_0to156"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Forest, houses? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a dense forest covering the landscape, with numerous houses visible nestled among the trees and along roads. The scene matches the described elements of forest and houses without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a forest forest with the lower, with houses houses scattered among among the trees. on the. The presence is the description ' of a and houses, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e81db939a9cb2b0e77efc78945bfff1207eca1b2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c470f2857367ea2f313ae4d50225cf2daf0559cd013a3c37418b184eed00157
+size 266976
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..3f35bc9c039419fad81a3a2712ce61666eb66131
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a woman in a green shirt discussing a developing story. The woman appears to be surprised or shocked by the news, as she is making a surprised facial expression and gesturing with her hands. The news segment is from the Good Morning Britain show, as indicated by the logo in the corner. The story being discussed is about John Lewis going gender-neutral, with the tag on their own brand items now reading 'Girls & Boys'. The background of the video shows a cityscape with a bridge, suggesting that the segment is being broadcast from a studio in a city. The style of the video is a typical news segment, with a focus on the reporter and the story being discussed."
+    ],
+    "video_ids": [
+        "5eUBgyxBQ04_1_863to993"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a green shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman wearing a green shirt, which matches the core description. The background and other elements do not contradict this observation.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features shows a woman wearing a green shirt, which matches the description description. The presence and additional elements, not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..aea3941ba2b5717bd93fe2f6eb5e7bdb46765e4c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1249-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e53b9280421f7f3c3e3b9a3fd6f2cbfe61055cbd5f10d0e50086095d4efbe3d
+size 215721
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..7dae65685f2eb3c8728f01c843bd1fe3ec7232f8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a car's engine bay, focusing on the various components and wiring. The style of the video is informative, likely aimed at providing a detailed look at the car's engine components for educational or diagnostic purposes. The engine bay is filled with a complex network of wires, hoses, and mechanical parts, all of which are interconnected and arranged in a specific manner. The colors of the components vary, with red, black, and silver being the most prominent. The video captures the intricate details of the engine bay, allowing viewers to observe the layout and design of the car's engine system."
+    ],
+    "video_ids": [
+        "6khbvV0SUy0_67_0to140"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Various components and wiring of the car's engine? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows various components and wiring of the car's engine, including hoses, wires, a pressure gauge, and metal framework. These elements are consistent with the description of engine components and wiring, even though a hand briefly appears in the frame. The core description is not contradicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts various components and wiring of a car's engine, including what, wires, and battery cap, and other parts. The elements are typical with the description of the components and wiring, and though the real is appears in the frame, The presence focus is fulfilled contradicted by}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..861012fbaa35ec3930ab9732de0b14dbdba48f9f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:465f97187f90c7ddbf11803d20f8c75409c1cb1d461d418446a22db473167458
+size 120933
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..7f867da838088561fe28d2ec48dbb7496a4b2f74
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen sitting in a cozy room with a green chair and a white brick wall in the background. She is wearing a pink shirt and a white apron adorned with red strawberries. A stuffed bunny is placed on a shelf behind her. The woman appears to be in deep thought or possibly daydreaming, as she gazes off to the side with her eyes closed. The room is filled with natural light, creating a warm and inviting atmosphere."
+    ],
+    "video_ids": [
+        "DmpAWm8b2pk_28_0to133"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a pink shirt, a white apron adorned with red strawberries, and a stuffed bunny.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman wearing a pink shirt and a white apron with red strawberries. A stuffed bunny is also visible in the background, sitting on a stack of books. All core elements described are present in the video without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a woman wearing a pink shirt and a white apron adorned red strawberry. There stuffed bunny is also visible on the background on placed on a shelf of shelves. The elements elements of in present in the video.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2e472f74df1883f20c90bd22dcd73c6347d18020
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c7b85f367676034e6bb3811fd19f29e1264b4e5c92afbedc9fd5e3d48a630ba4
+size 90328
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..f939c501356631eccf2c8cf819dfd3c867a9198d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a small, transparent aquarium placed on a wooden surface. The aquarium is elevated by white plastic bottles, creating a makeshift stand. Inside the tank, there are several decorative elements: black rock formations resembling mountains, a small red Ferris wheel, and a green structure that looks like a miniature house or temple. A few orange fish swim gracefully around these decorations, adding life to the scene. The water level is high, almost reaching the top of the tank, and bubbles can be seen rising from the bottom, indicating an active filtration system. The background is dark, which makes the vibrant colors of the fish and decorations stand out prominently. The video captures the serene movement of the fish and the gentle flow of the water, providing a peaceful and visually appealing scene. There are no significant changes or movements throughout the video; it maintains a steady focus on the aquarium and its inhabitants."
+    ],
+    "video_ids": [
+        "083591918c7a00d530e53d3a5f1d2e1a35ed98ce9be1baa7c44fb6e9c24f7d37"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Black rock formations, a small red Ferris wheel, a green structure (miniature house or temple), orange fish, bubbles, and water.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing black rock formations, a small red Ferris wheel (or similar red circular structure), a green structure resembling a miniature house or temple, orange fish swimming, bubbles rising, and water filling the tank. All specified elements are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a rock formations, a small red Ferris wheel,which similar structure structure object), a green structure that a miniature house or temple, orange fish,, bubbles rising from and water filling the tank. The these elements are present and match with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..da2cf607c597c1cb58f512e60e6d1a8aa0aca6c9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f2592426e48fdb3755e4b6377cb9098a7b618698702b5caf0bdf4cd28e7b6e2
+size 159036
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..7366e20fce42d07e29b0950c4082a65e73bb533b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a black Mercedes-Benz car driving on a snow-covered road. The car is sleek and modern, with a shiny black exterior that reflects the surrounding environment. The road is covered in a layer of snow, and the car's tires leave tracks as it moves forward. The car is driving towards the camera, and the angle of the shot suggests that the car is moving at a moderate speed. The background is a mix of trees and houses, indicating that the car is in a suburban or rural area. The overall style of the video is realistic and naturalistic, with a focus on the car and its movement through the snowy landscape."
+    ],
+    "video_ids": [
+        "FlO_6Cy8EFA_6_105to233"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black Mercedes-Benz car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black Mercedes-Benz car parked on a snowy surface, which matches the description. The car's design, emblem, and color are consistent with a Mercedes-Benz model, and there are no elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black Mercedes-Benz car parked on a snowy driveway. which matches the description of The car is design, including, and overall are consistent with the Mercedes-Benz.. and the are no additional in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5717c91a3af9644ed7701c2c17aecb57c55777fd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:162a62bbe56d05a25350b11b24702c9566048ab3d18a735490f474189ea5e98f
+size 293479
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..b75b195bd1d295f28a17ce76aeb7d63e3e4e1532
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a stylish and modern restaurant interior. The space is bathed in a cool blue light, creating a sophisticated ambiance. The main focus is a long, white table with a sleek design, surrounded by comfortable blue chairs. The table is adorned with elegant blue glassware, adding a touch of elegance to the setting. The room is filled with various pieces of furniture, including additional chairs and tables, creating a cozy and inviting atmosphere. The walls are decorated with tasteful artwork, enhancing the overall aesthetic of the space. The video showcases the restaurant's design and layout, highlighting its modern and stylish elements."
+    ],
+    "video_ids": [
+        "WxrBNVBv8Lw_33_0to159"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Long, white table, comfortable blue chairs, elegant blue glassware, various furniture (chairs, tables), tasteful artwork.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a long white table surrounded by comfortable blue chairs. Elegant blue glassware is visible on the table, and various furniture pieces including chairs and tables are present throughout the space. Additionally, tasteful artwork, including framed pieces on the walls, is displayed. All these elements align with the description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as showcasing showing a long, table, by comfortable blue chairs. The blue glassware is also on the table, and there pieces,, chairs and tables are present. the scene. The, thereeful artwork is such a pieces on the walls, is visible, The elements elements align with the description provided any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..15bde8e133ada60f1b90ef837c3de70c9d21c82e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e10aa8082c5b033e032d8926b5a6d5e45594c54e203757ced5aad529d519f2dd
+size 115368
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..0ffacf4617bafcd31fb06c6e04b23c93be4e3c16
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young woman with blonde hair is seen in three different frames. She is wearing a straw hat and a pink top. In the first frame, she is standing and looking to her left with her mouth open, as if she is surprised or in the middle of a conversation. In the second frame, she is still standing but has turned her head to the right, looking at someone or something off-camera. In the third frame, she is sitting down at a table with her hands clasped together, appearing to be engaged in a conversation or listening intently. The background of the video is blurred, but it seems to be an indoor setting with a white wall. The style of the video is casual and candid, capturing the woman in a natural and unposed manner."
+    ],
+    "video_ids": [
+        "M8h_609KPsA_17_21to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman with blonde hair wearing a straw hat and a pink top.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman with blonde hair wearing a straw hat and a pink top, which matches the description. Although there are other people in the scene, the main subject clearly fulfills the specified criteria.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young woman with blonde hair wearing a straw hat and a pink top. which matches the description provided The the are two elements in the video, the main focus consistently fitsfills the ' '.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0519a6d68ffcd59a8550b1db2e7123c2fe7ffac1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:27e8f996493f1a36d6b16f596ee2b27286064704323108e079a4497ab4bf0bac
+size 148043
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..e175859f486bb7415fde61efbd03efd1a8540151
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a panel of four people discussing a developing story. The panel consists of two men and two women, all dressed in business attire. They are seated around a curved desk with microphones in front of them. The background is a large screen displaying a cityscape, and there are various logos and social media icons at the bottom of the screen. The text overlay on the screen reads \"TRUMP TAKES AIM AT MAY\" and \"US President says we will no longer deal with Sir Kim Darroch\". The style of the video is a typical news broadcast with a focus on current events and political discussions."
+    ],
+    "video_ids": [
+        "L-HLeTXFME4_1_28to210"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A panel of four people (two men and two women), a curved desk, microphones, a large screen, logos, and social media icons.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a panel of four people (two men and two women) seated around a curved desk. Microphones are visible on the desk, and there is a large screen displaying a cityscape in the background. Logos for 'Good Morning Britain' and 'ITV' are present, along with social media icons (Twitter, Facebook) on the screen. All elements described in the condition are present, and no contradictions are found.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a panel of four people,two men and two women) seated at a curved desk. Therephones are visible in the desk, and there is a large screen in a cityscape in the background. Logos and variousTR Morning Dubai' and socialTheV' are present, along with social media icons atFacebook, Facebook, at the screen. The elements match in the question are present in making the contradictions are noted.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..746c99e91495f507f0385bf72463247f5c0838f0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:56125c9176f2ffa37cc8ef480551e8507a3780ce6e8a1fe0bafe6c0a1fdb6407
+size 124061
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..dbe6b9ec58e487bb4207da0f4c056e3e64ee76c9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a green engine with a silver fan in the center. The engine is shiny and appears to be made of metal. The fan has a series of blades that are evenly spaced. The engine is mounted on a black base, which is also made of metal. The engine is the main focus of the video, and there are no other objects in the frame. The style of the video is a close-up shot, which allows for a detailed view of the engine and its components. The video does not contain any text or additional elements. The engine is stationary, and there is no movement or action in the video. The video is simple and straightforward, focusing solely on the engine and its design."
+    ],
+    "video_ids": [
+        "XvzLJUnVh_g_2_0to158"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A green engine with a silver fan and a black base.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a green turbocharger with a silver turbine wheel (fan) and a black base (central hub). The object matches the description of a green engine with a silver fan and black base, even though it is a turbocharger rather than a traditional engine. The core description is accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a green enginecharger with a silver fan and andfan) and a black base,housing housing). The green matches the description of a green engine with a silver fan and a base, fulfilling though the is not turbocharger, than a traditional engine. The presence elements is accurately represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cac997a5e5e034f23fb6801b1da185d997c141bf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-125-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fc5573a641ebe9f81a0beceb27dc1cb50e74d78e6388a1c0426352a6d0a6f81b
+size 292884
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..977e2b75a62b67840945a6b4374836d27127be97
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse of a garage being organized. It starts with a cluttered garage filled with various items, including a bicycle, a red gas can, and a white bucket. As the video progresses, the garage is transformed into a well-organized space with items neatly arranged on shelves and in storage bins. The final frame shows a clean and tidy garage with all items in their designated places. The style of the video is a simple, straightforward time-lapse that captures the transformation of the garage from cluttered to organized."
+    ],
+    "video_ids": [
+        "9Ws3XqJWs_I_4_393to515"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: bicycle, red gas can, white bucket, shelves, storage bins? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a garage setting with a motorcycle (not a bicycle) in the background, red gas cans, a white bucket, shelves, and storage bins, which largely matches the described elements. The presence of a motorcycle instead of a bicycle does not contradict the core description, as it is a common variation in such settings.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a storage with with a bicycle,which a bicycle), and the foreground, a gas cans, white white bucket, and, and storage bins. which align matches the description objects. The presence of a motorcycle instead of a bicycle is not contradict the core description, as the is a similar object in garage settings.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..96e02412d9b8fece3c1153aa2d150dc8c7a51b43
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0cffb73191b2837ccbe7a27ad8b7f89f2d9136640d8cefb8e5e53a3348b57a1c
+size 246499
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..ffdb0f9a4204f513296946e4975565d528cae80a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a luxury sports car, focusing on the steering wheel and dashboard. The steering wheel is black with a silver emblem in the center, and the dashboard is sleek and modern, featuring a digital display. The car's interior is well-lit, highlighting the high-quality materials and craftsmanship. The video is shot from the perspective of the driver, providing a sense of the car's design and features. The style of the video is sleek and professional, with a focus on the car's interior and design elements."
+    ],
+    "video_ids": [
+        "XVkOpLoLfHc_38_0to102"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, dashboard? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a steering wheel and dashboard as the central focus, with the steering wheel prominently displayed in the foreground and the dashboard visible behind it. The details of both elements are consistent with the description, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a steering wheel and a, described main elements. which the steering wheel prominently displayed in the foreground and the dashboard visible in it. The presence of the the, consistent with the description, and there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fa117b887d13fa1defe0305b22dd99a09ed87bb9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:704ebb5cd79f56960c841a6d3971be4f484b0652e820753077c07c7870d84d58
+size 196915
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..b51c26606a78408212e613ab7903f772cd3382f6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a baseball player in action, showcasing his skills and enthusiasm for the game. The player, dressed in a white and blue uniform with the word \"Rays\" emblazoned across his chest, is seen in three distinct moments.  In the first frame, the player is seen in a dynamic pose, his body coiled in anticipation of the pitch. His black helmet gleams under the stadium lights, reflecting his focus and determination.  The second frame captures the player in the midst of a powerful swing. His bat, gripped firmly in his hands, is poised to connect with the incoming ball. The intensity of the moment is palpable, as the player's muscles tense and his eyes follow the trajectory of the ball.  The third frame shows the player in a triumphant pose, his body language radiating joy and satisfaction. His smile is wide and genuine, reflecting the thrill of a successful hit. The player's gloves, worn on his hands, are a testament to his dedication to the sport.  Throughout the video, the player's uniform remains consistent, with the \"Rays\" logo standing out prominently. The background, though blurred, gives the impression of a bustling baseball stadium, filled with the anticipation and excitement of the game. The video is a dynamic and engaging portrayal of a baseball player's journey from anticipation to action to celebration."
+    ],
+    "video_ids": [
+        "BJo4YgBgl8s_15_139to270"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Baseball player, baseball, bat, gloves? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a baseball player wearing a Rays uniform, gloves, and a helmet, which matches the 'Object(s)' condition. Although the baseball and bat are not visible in the frame, the player's attire and context strongly imply he is a baseball player, and the presence of gloves is confirmed. The absence of the baseball and bat does not contradict the core description, as they are not required to be visible for the condition to be met.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a baseball player in a uniform uniform, holding, and holding bat, holding are the 'Base(s)' condition. The the video is bat are not visible in the frames, the player's attire and equipment strongly suggest the is a baseball player, fulfilling the presence of gloves and evident. The video of the baseball and bat does not contradict the ' description, as the are not required to be visible in the ' to be met.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1463c8b55103c87e70b86f71abfd3bc5e89263be
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68cb6a9068a1f5de8bca6b013890da8c4bfa06fdaa4f2b6e8f340bc6401fbbab
+size 209008
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..e39a498eb711306d080a25309382752a70eeb11f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse of a car's interior, showcasing the design and features of the vehicle. The car is parked in a forested area, with the interior bathed in natural light. The steering wheel is prominently displayed in the first frame, with the car's logo in the center. In the second frame, the car's dashboard is visible, showcasing the infotainment system and other controls. The third frame shows the car's door, revealing the door handle and window controls. The car's interior is designed with a focus on comfort and luxury, with high-quality materials and a sleek design. The time-lapse style of the video allows viewers to see the car's interior in a dynamic and engaging way."
+    ],
+    "video_ids": [
+        "L9fdNMAo97g_11_0to131"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, car's logo, dashboard, infotainment system, controls, car's door, door handle, window controls? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows the steering wheel, car's logo (Buick emblem on the steering wheel), dashboard, infotainment system (though not fully visible, the central console suggests it), controls (buttons on the steering wheel and dashboard), car's door (visible on the right side), door handle (visible on the door panel), and window controls (visible on the door panel). All requested elements are present and clearly visible.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful the steering wheel, car's logo,partick),), the steering wheel), dashboard, infotainment system,touch the in visible), the screen screen and it), controls (ste and the steering wheel and dashboard), car's door,inter in the right side), door handle (part on the door),), and window controls (part on the door panel). The these objects are present and identifiable identifiable in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..784188763162bfc6a87fa32986f70bb8979a7ba3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9fb5293cb2ee3bbbb1debe8e61cb0f3e9511e5b99df4f849ee70f2c419d27b47
+size 193774
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..3164ce89eb071f6e4818e86874c249fb0550c166
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the life cycle of a plant, from its vibrant green leaves to its eventual wilting and decay. The plant starts off with lush green leaves, full of life and vitality. As the video progresses, the leaves begin to wilt and turn brown, signifying the plant's decline. The leaves continue to wilt and eventually fall off, leaving the plant bare and lifeless. The video ends with the plant's stem and branches, devoid of any leaves, standing stark against the backdrop. The entire process is a poignant reminder of the cycle of life and the inevitable decline of all living things."
+    ],
+    "video_ids": [
+        "zXO_j0vriwk_17_18to233"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A plant with leaves? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a plant with leaves, some of which are green and healthy while others are brown and wilted. This matches the description of a plant with leaves, even though the condition of the leaves varies.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a plant with leaves, which of which are large and others, others are brown and appeared. The align the description of a plant with leaves, even though the leaves of some leaves varies.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c3dadb2304be25ba7bf2b9818beb4a26b488a132
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e869a6810d646ca27aa0976e9d5b09df751a805b82344a77225dbd476c1b6ed0
+size 249260
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..37cf3b16a154d0991a3becc3c9e5e57a1814672c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a suit standing in front of a microphone, presumably at a press conference or event. He is wearing a blue suit and a tie, and appears to be speaking or listening intently. Behind him, there is a woman standing with her arms crossed, possibly waiting for her turn to speak or listening to the man. The background features a wall with the words \"Jameson Empire Awards\" repeated multiple times, indicating that the event is likely related to the Jameson Empire Awards. The style of the video is straightforward and documentary-like, capturing the man and the woman in a candid moment at the event."
+    ],
+    "video_ids": [
+        "NFNC4LBFCdg_12_0to151"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue suit and tie, a woman with her arms crossed? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a blue suit and tie speaking into a microphone, and a woman standing behind him with her arms crossed. These elements are consistent with the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man in a blue suit and tie standing at a microphone, and a woman standing behind him with her arms crossed. The elements match consistent with the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bb0afe03c95cddb0d00f9e5ffa76b3c7287f4ba9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dbb341a80a53a6aad239674731ea0f54e97f4b27fcfbc12901e8d03a281d00b5
+size 152707
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..dd242d4c93e141870590c847512b7f6f3364241f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a vibrant and colorful depiction of a bouquet of flowers, specifically roses and lilies, set against a blue sky background. The flowers are arranged in a way that they appear to be floating or floating in the air. The roses are in full bloom, with their petals spread out, while the lilies are in various stages of bloom, with some fully open and others still closed. The colors of the flowers are vivid, with shades of pink, purple, and white. The leaves of the flowers are green and appear to be in good condition. The overall style of the video is realistic, with attention to detail in the depiction of the flowers and their surroundings. The video does not contain any text or other objects. The focus is solely on the flowers and their beauty."
+    ],
+    "video_ids": [
+        "6zx0FImKtEY_29_0to144"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bouquet of roses and lilies, flowers in various stages of bloom, green leaves? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video displays a vibrant floral artwork featuring roses and lilies in various stages of bloom, accompanied by green leaves. The composition includes multiple flowers with detailed petals and stamens, matching the description of a bouquet with flowers in different bloom stages and foliage.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a bouquet bouquet arrangement featuring a and lilies in various stages of bloom, with by green leaves. The presence includes a roses, different petals and leavesamens, which the description of a bouquet with flowers in different stages stages. green.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6853aac47a85017f9c032c8dcc9017bc8789dd8d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d5ddcddaf09b103644e31f9e14635ddf3ea4764479352b6b20845d2d990a365
+size 114435
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..2f4ba559040edcd53a02508993fb1001a965356e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person using a food processor. The food processor has a stainless steel bowl and a black lid. The person is holding the lid and appears to be in the process of either removing it or placing it on the food processor. The background is blurred, but it seems to be a kitchen setting. The style of the video is a simple, straightforward demonstration of how to use a food processor. The focus is on the food processor and the person's hands, with no additional context or embellishments."
+    ],
+    "video_ids": [
+        "_a9CN5rNT2w_7_84to232"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person, food processor (stainless steel bowl, black lid), lid.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person interacting with a stainless steel bowl that is part of a food processor or pressure cooker, which has a black lid. The person's hands are visible, and the lid is being held open, matching the description. The core elements \u2014 person, stainless steel bowl, and black lid \u2014 are clearly present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person interacting with a food steel food that appears part of a food processor. similar cooker. which has a black lid. The person is hands are seen, and they lid is being removed and, which the description of The presence elements of the, food steel bowl, black lid lid \u2014 are present present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8a1b531b285ec99e91e18b8994f2c1246fc9a1cf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1250-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:94938a835a63540cdf3053a215bb5c3feeaad725014364b92899a43a98b737e5
+size 108364
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..a6979db66d85e62df9b29849029f59db3b6befce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a television show segment featuring a woman sitting on a set with a bookshelf in the background. The woman is wearing a black dress with a sequined pattern and has curly hair. She is speaking and gesturing with her hands. On the screen, there is a poll question that reads, \"Are You afraid of Barbara Walters?\" with the results showing 16% for \"Yes\" and 84% for \"No.\" The poll is from a website called \"www.whl.tv.\" The style of the video is a standard television interview or talk show segment, with a focus on the poll question and the woman's reaction to it."
+    ],
+    "video_ids": [
+        "NJj-18em32I_11_294to462"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a bookshelf, a poll question on the screen.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman seated in front of a bookshelf, and a poll question about fear of Barbara Walters is prominently displayed on the screen. These elements are consistent with the 'Object(s)' condition described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features shows a woman sitting in front of a bookshelf, which a poll question appears ' of heights Eden is displayed displayed on the screen. The elements match consistent with the 'Object(s)' condition provided.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ac712afd17fd70864d5f8b6cf2fb020899a8e10f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:587c2a88bceed7b38e092566deb43cb450ac22b15f7439c6668bef1a0fe367ff
+size 161180
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..447e5e4735dd26248cfff4adf071dc3571a9ba93
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a sleek, modern sports car parked in front of a large garage with a stone facade. The car is painted in a combination of dark and light colors, with a prominent yellow stripe running along its side. The car's design is sleek and aerodynamic, with a low and wide stance that suggests speed and power. The garage door is open, revealing a spacious interior that is well-lit and clean. The car is parked on a concrete surface, and the garage is situated in a quiet, secluded area. The overall style of the video is realistic and detailed, with a focus on the car and its surroundings. The lighting is natural and even, highlighting the car's design and the garage's architecture. The video does not contain any text or additional elements, and the focus is solely on the car and its immediate surroundings."
+    ],
+    "video_ids": [
+        "P_7piye1Who_15_0to120"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Sleek, modern sports car with a dark and light color combination and a prominent yellow stripe, parked on a concrete surface.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a sleek, modern sports car with a dark gray body and a prominent yellow stripe running along the hood and side. The car is parked on a concrete surface in front of a stone building with arched doorways. The description of the car's color combination and stripe is accurately reflected in the video, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a sleek, modern sports car with a dark and and and a prominent yellow stripe. along the side and sides. The car is parked on a concrete surface in front of a garage garage with ached windowsways, The car matches the car's appearance combination and the is accurately represented in the video.\"\n fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..117a623816949501a43958765d65c2f715b93f31
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:31911d03adc3289d9aff8bdf33db7d9ed1f860e6b153c809669af5ae029000a5
+size 200796
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..310a4e8d71664003126e325ad5766be93341f48a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a vibrant green sports car driving on a paved road. The car is sleek and modern, with black rims and a shiny exterior. The car is moving from left to right across the frame. In the background, there are people standing on the side of the road, watching the car pass by. The sky is clear and blue, suggesting a sunny day. The road appears to be in a rural or suburban area, with trees and bushes lining the sides. The overall style of the video is dynamic and energetic, capturing the motion of the car and the reactions of the onlookers."
+    ],
+    "video_ids": [
+        "RXv_xIQ2eig_14_28to191"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A vibrant green sports car, people standing on the side of the road.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a vibrant green sports car, specifically an Audi Q8, parked on a paved surface. Several people are visible standing around the car, some taking photos or videos, which matches the description of people standing on the side of the road. The core elements of the description are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a vibrant green sports car, which a Aston R8, which on the road road. There people are visible standing on the car, which on photos, observing, which ful the description of people standing on the side of the road. The scene elements of the description are present represented in}<|im_end|>\nGuidId",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c5f60f34fa29acc4e39af26d73e5a16831d005dd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc3feadcfb68e4eedf1d5e3df09f158f8ecafdae2d91122d1ac550b9f206ee68
+size 426491
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..dbd4f309b805ec6a8d2cb15608602d8d34167e8f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and mustache, wearing a Hawaiian shirt with a floral pattern. He is seated in front of a brick wall, which serves as the backdrop for the scene. The man appears to be engaged in a conversation, as suggested by his open mouth and attentive gaze. The overall style of the video is casual and relaxed, with the man's attire and the brick wall setting contributing to this laid-back atmosphere. The video seems to capture a candid moment, perhaps during a casual interview or a friendly chat."
+    ],
+    "video_ids": [
+        "QXyjZJ0Ym64_41_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man with a beard and mustache, wearing a Hawaiian shirt with a floral pattern.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and mustache, wearing a Hawaiian shirt with a floral pattern. The description matches the visual content accurately, with no contradictions. Additional elements like the brick wall and bookshelf in the background do not conflict with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and mustache, wearing a Hawaiian shirt with a floral pattern. The man matches the visual elements of, with no additional or The elements such the brick wall in the in in the background do not conflict with the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ecf6235e6d7c6edb048233ead0709d5dd16f8914
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f121e163807f7f75b7785c711160a4a611715eb2a75ea231f31dae59968b118
+size 214965
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..aaaccc8c46849007b243bb566814c2971c11be64
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a woman and a man. The woman is standing in front of a green screen with a cityscape and a sports ball on a shelf in the background. She is wearing a black and white patterned dress and is speaking into a microphone. The man is shown in a separate frame, wearing sunglasses and a gold chain necklace, and is also speaking into a microphone. The style of the video is a typical news report with a focus on the two individuals speaking."
+    ],
+    "video_ids": [
+        "4G_pPI13QH0_5_0to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman and a man? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman in a studio setting and a man (in two different contexts: as a boxer and as a rapper in an inset image). Both individuals are prominently featured, fulfilling the 'Object(s): A woman and a man' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a woman and the black setting, later man appearsw sunglasses different frames) one a separate and as a person) a animated video) The are are present featured, fulfilling the 'Object(s)' A woman and a man' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5538d725c6f570a0d1aaea537a640db1104177e3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d061010909730df83fbf7497348a5756bde1b04f76c281f4c4f7df29c7014793
+size 307339
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..0fe6cf0a4b8d70ba0510fee5661cec3deb649df7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is on two players, one in a blue and orange uniform and the other in a red and white uniform. The player in blue and orange is in the process of being tackled by the player in red and white. The player in blue and orange is on the ground, holding onto the football, while the player in red and white is on top of him, attempting to bring him down. The action is intense and the players are fully engaged in the game. The background is a blur of green, indicating that the game is taking place on a grassy field. The style of the video is realistic and captures the intensity and excitement of the sport."
+    ],
+    "video_ids": [
+        "vJqcpElzW8o_48_0to104"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two players - one in blue and orange, the other in red and white uniforms.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two football players engaged in a tackle. One player is wearing a blue and orange uniform (Chicago Bears), and the other is wearing a red and white uniform (Washington Redskins). The uniforms match the description, and there are no conflicting elements that contradict the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts two players players, in a tackle. One player is wearing a blue and orange uniform,likely Bears colors and the other is wearing a red and white uniform (likely Command). The presence and the description provided and the are no additional elements in contradict the coreObject(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..54fb1140573f94c3e657dda19fef7c543efb2863
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:35996b4b6afcd05544ac2aaa555d44a3f5d0161975e60fa9bc0e63adce807c2d
+size 299887
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..2cc40630e0bed9b6a2772daf1d8ec373d8b88da6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the elegance and speed of a red sports car in motion. The car, with its sleek design and shiny exterior, is the main focus of the video. The car is seen from a side angle, showcasing its aerodynamic shape and the details of its design. The car's headlights are on, illuminating the road ahead and adding to the sense of motion. The car's wheels are also visible, emphasizing its powerful stance. The background is blurred, indicating the car's high speed and the focus on the car itself. The overall style of the video is dynamic and exciting, capturing the thrill of driving a high-performance sports car."
+    ],
+    "video_ids": [
+        "3AEXlCdEqE4_32_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red sports car, specifically a Ferrari, with its distinctive design, headlights, and wheels visible. The car is the central focus and matches the description of a red sports car.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a red sports car, which a Ferrari, which a distinctive design features including, and wheels.. The car is in central focus of matches the description of a red sports car.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7a19eacf88d4dcd7c0b46eb8177473a3fe49afb1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:39c413ef6e4e098866f01c1cac93c90088c595ad8bff72f15f5fdf2cee489376
+size 237373
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..3a471bc3ca8e97c5c85fdd15723b544a672e5bbb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two men are seen in an office setting. The man on the left, wearing glasses and a blue shirt, is seen smiling and gesturing with his hands, possibly explaining something or engaging in a conversation. The man on the right, dressed in a suit and bow tie, is holding a coffee cup and appears to be listening attentively. The office environment includes shelves filled with various items and a whiteboard in the background. The overall style of the video suggests a casual and friendly work atmosphere."
+    ],
+    "video_ids": [
+        "57OVtixRzIw_8_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men - one on the left wearing glasses and a blue shirt, the other on the right in a suit and bow tie.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men as described: one on the left wearing glasses and a blue shirt, and another on the right dressed in a suit with a bow tie. The visual details match the description accurately, with no contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men in described in one on the left wearing glasses and a blue shirt, and the on the right in in a suit and a bow tie. The setting elements match the description,, and no additional or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..44d33aa72905a08678bb834fe9be4c7db78eb448
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1251-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3712432a8e8c9c650b44acee91e7c745a782f58fbabc62aadd7e6d94800844e
+size 134786
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..328fb0e91d003da3652968f7e518a6a18a31e677
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a stack of cookies on a plate, with a focus on the textures and colors of the cookies. The cookies are golden brown and have a crumbly texture, with visible chunks of nuts and seeds. The plate is placed on a reflective surface, which adds a glossy sheen to the cookies and the plate. In the background, there are two cups of coffee, suggesting a cozy and relaxed setting. The style of the video is simple and straightforward, with a focus on the food and the textures, rather than any action or movement. The lighting is soft and warm, highlighting the golden brown color of the cookies and the reflective surface of the plate. The overall impression is one of comfort and indulgence, with the cookies as the main subject of the video."
+    ],
+    "video_ids": [
+        "-umPwgYGitM_25_0to148"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A stack of golden brown, crumbly cookies with visible nuts and seeds, two cups of coffee.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a stack of golden brown, crumbly cookies with visible nuts and seeds on a plate, which matches the core description. Additionally, two white cups (likely for coffee or tea) are visible in the background, fulfilling the 'two cups of coffee' part of the description. The presence of a baking tray with more cookies and scattered nuts does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a stack of golden brown cookies crumbly cookies with visible nuts and seeds, a white. which matches the description description. Additionally, there cups cups arelikely containing coffee) tea) are visible in the background, fulfilling the 'two cups of coffee' condition of the description. The presence of the wooden tray and a cookies and a nuts in not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..47c43d71a42b1b5692f7d4e56c43ba0acc4783cc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e01e4d01b599033aa575f9c567b66ac8eec30fa82dea15852265b3dbce1e078f
+size 94421
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..f8b8128e06348219315884b88e78742f0b2cdd0d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a stylized, animated depiction of the superhero Captain America. He is shown in three frames, each capturing a different pose and action. In the first frame, he is seen holding a shield with the American flag design, standing confidently with his arms crossed. In the second frame, he is shown in a dynamic pose, running forward with his shield held in front of him. In the third frame, he is seen in a fighting stance, ready to strike with his shield. The background of each frame is a vibrant, red, white, and blue American flag, emphasizing the patriotic theme of the character. The style of the video is reminiscent of comic book art, with bold lines and vivid colors."
+    ],
+    "video_ids": [
+        "NJB_n7XrPNc_18_18to198"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Captain America, shield? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features Captain America in his iconic costume, holding his signature shield. The visual elements are consistent with the description, and no conflicting elements are present that would contradict the core description of Captain America and his shield.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a America, two iconic costume, holding his shield shield. The shield elements, consistent with the depiction, showing the additional elements are present. would contradict the ' description of the America and his shield.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bd6bcdb134f4f3cdcb6fba8696a7238b110588bd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af455a1f5295b92ab156fc00dbc4cf46c1e010e4f58ef75f789cff360aefc083
+size 298567
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..cbd70cb2249be5c112836e718e826b6a73486e17
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a stylized, animated depiction of a character in a futuristic setting. The character is wearing a red and black suit with a helmet and is holding a large gun. The character is shown in three different poses, each with a different expression and stance. The background features a fiery explosion and a blurred figure in the background, suggesting a battle or conflict. The overall style of the video is dynamic and action-packed, with a focus on the character's movements and the intensity of the scene."
+    ],
+    "video_ids": [
+        "0_TaTiu5ssw_94_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A character wearing a red and black suit with a helmet and holding a large gun.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The central character in the video is wearing a red and black suit with a helmet and is holding a large gun, which matches the description. The background elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video character in the video is wearing a red and black suit with a helmet, is holding a large gun. which matches the description provided The background elements, not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c6f9e7dd3ab4a8fa7a90c7cd5b969921e5fb3c8d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f93696992acca510095870ddd79b35e24e178eabb5b9e5a67b47bfe4c70055f0
+size 208710
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..a7cd9f86b686deabfa655c8dadd6bd6f844233b8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man dressed in a black robe and a white shirt, standing in a room with a wooden door and a window with blinds. He is wearing glasses and a mustache. The man is looking to the side, and his expression is serious. The room has a lamp in the background. The style of the video is realistic, and it appears to be a still image rather than a moving one."
+    ],
+    "video_ids": [
+        "aXyDAi51A7s_14_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man dressed in a black robe and a white shirt, wearing glasses and a mustache. There is also a lamp in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed in a black robe and white shirt, wearing glasses and a mustache, which matches the core description. A lamp is also visible in the background, fulfilling that part of the description. There are no conflicting elements that contradict the given details.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed in a black robe and a shirt, wearing glasses and a mustache, which matches the description description. The lamp is also visible in the background, fulfilling the part of the '. The are no additional elements in contradict the given description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b65b587be007f202416ec5abf542b1df6ac0d8fa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc6c69aa88c3cd42f0c8ab1f34b29baf0316c9d4699da214818e4658bf220935
+size 67583
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..15437883dc8003ba9cfe198b56c68ceda33fa69c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a 3D animated character, a young man with a surprised expression. He has short brown hair and is wearing a red jacket with a hood. The character is standing in front of a pink, textured wall. In the first frame, the character is looking to the left with his mouth slightly open. In the second frame, he is pointing his finger to the left. In the third frame, he is holding a blue object in his right hand. The style of the video is cartoonish and colorful, with a focus on the character's exaggerated facial expressions and body language."
+    ],
+    "video_ids": [
+        "EB_Ov4FSNoM_1_0to164"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man with short brown hair and a red jacket with a hood, a blue object.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man with short brown hair, wearing a red jacket with a hood, and he is holding a blue object (which appears to be a trekking pole or similar item). The description matches the core elements of the character and his attire, and there are no contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young man with short brown hair wearing wearing a red jacket with a hood. which a is interacting a blue object ina appears to be a phoneking pole or similar item). The video matches the core elements of the video and the attire, and the are no contradictions with}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0e2f1e521fc3e6fbc8c56af8456eaf462e4e3ff2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5aa9872a1d64f867a6fbc76e366f419a5f5a0814f00833fe13a15467ae2d4627
+size 203671
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..f8e87fb046c0181a25c25f784880c794ce70b66a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man with a bald head and a beard is seen in a living room setting. He is wearing a white t-shirt and gray pants. In the first frame, he is holding a wooden object, possibly a bat or a stick, in his right hand. In the second frame, he is swinging the object with force, indicating that he is in the middle of an action. In the third frame, he is still holding the object, but the action seems to have stopped. The living room has a couch and a bookshelf in the background. The man's expression and the motion of the object suggest that he is engaged in a playful or energetic activity."
+    ],
+    "video_ids": [
+        "3HxKLx45oCY_47_0to117"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a wooden object (possibly a bat or a stick).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man holding and swinging a wooden object connected by a chain, which resembles a nunchaku. The object is clearly visible and matches the description of a wooden bat or stick. No elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a man holding a swinging a wooden object that to a black, which resembles a batunchaku. The object is indeed wooden and fits the description of a wooden bat or stick. The other contradict the description description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ac45d0b3224eb3bca2f4b5c74ff7f5ad81a4f7f9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d1b4355837293ddfc5878ce7d6446bfa7b5c5e1d4af91232e903bc79747a9a3f
+size 157986
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..779260b1cf3c6b665ca7da57024d15b031f24269
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a luxurious boat interior, featuring a spacious living area with a curved sofa and a coffee table. The sofa is adorned with multiple cushions, and the coffee table holds a vase with flowers. The floor is covered with a plush rug, and the walls are decorated with a combination of wood and metal panels. The lighting is soft and ambient, creating a warm and inviting atmosphere. The style of the video is a blend of interior design and travel, highlighting the comfort and elegance of the boat's living space."
+    ],
+    "video_ids": [
+        "0yCmPEGnubk_30_0to130"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: curved sofa, coffee table, cushions, vase with flowers, plush rug, wood and metal panels on the walls? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a curved sofa with multiple cushions, a central coffee table with a vase containing flowers, a plush rug underneath, and walls featuring wood and metal panels. All specified elements are present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a curved sofa with multiple cushions, a coffee coffee table, a vase containing flowers, and plush rug on, and wood with wood and metal panels. The the elements are present and match depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e7a859e88c3c4a93d0ce451bbe3082dc911dc397
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:18c0a348abc7846fbd7ea4eaad88aa2c4ff139f771916977f77f7eb7019fab02
+size 196973
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..69b8979170d926723d383fdd27f0a20baa925612
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a brown and white pit bull dog on a leash, standing in a grassy field. The dog is wearing a blue collar with a silver leash attachment. The dog appears to be panting and looking to the left. The background features a blurred green field and trees, suggesting an outdoor setting. The style of the video is a simple, straightforward capture of the dog on a leash, with no additional action or movement. The focus is on the dog and its immediate surroundings."
+    ],
+    "video_ids": [
+        "_IhsbhPR4f4_12_246to401"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A brown and white pit bull dog on a leash, wearing a blue collar with a silver leash attachment, appearing to be panting and looking to the left.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a brown and white pit bull dog on a leash, wearing a blue collar with a silver leash attachment. The dog appears to be panting with its mouth open and is looking to the left. All core elements of the description are accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a brown and white pit bull dog on a leash, wearing a blue collar with a silver leash attachment. The dog appears to be panting and its mouth open and is looking to the left. The these elements of the description are present represented in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..27f374bfe28b78817af76cb82fe3e41eb1b807c0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1252-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1032fa5afd0c478a070e838fcff25779a89803fa7a22df8d7d572e5dcb42275
+size 169133
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..53ea023ef863235b02c7f8af34ea86d6003d5b59
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a bald man wearing glasses and a brown jacket standing in front of a large, old stone house with a black car parked in front of it. The man is looking directly at the camera, and the house has a classic, historic design with multiple windows and a prominent entrance. The sky is clear and blue, suggesting a sunny day. The man appears to be posing for the photo, and the overall style of the video is casual and straightforward, with no additional action or movement."
+    ],
+    "video_ids": [
+        "Ff1yRQKI3nA_0_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bald man wearing glasses and a brown jacket? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a bald man wearing glasses and a brown jacket, which matches the description. The presence of a woman beside him does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a bald man wearing glasses and a brown jacket. which matches the description provided The background of the building in him and not contradict the core description of}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5d43a512ee0a5d6a911985b157596c5edafb7549
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2085053822f518c205072986b3792bc70af283e1d67e6f37188135d4098e547
+size 175365
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..69782bafda5140273b6645c5c690c50fa8ccc5e4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red toolbox with various automotive fluids and tools on top of it. The fluids include two bottles of blue Peak antifreeze, a white bottle of water, and a red bottle of oil. A yellow funnel is also present on the toolbox. In the background, there are three cars with their hoods open, indicating that they are being worked on. The style of the video is a straightforward, real-life depiction of a car repair or maintenance scenario. The focus is on the toolbox and the fluids, with the cars in the background serving as context for the activity. The video does not contain any stylized elements or artistic effects."
+    ],
+    "video_ids": [
+        "29z94xBdAjE_13_853to980"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red toolbox, two blue Peak antifreeze bottles, one white water bottle, one red oil bottle, one yellow funnel, three cars with open hoods? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a red toolbox with various items on top, including two blue Peak antifreeze bottles, one white water bottle, one yellow funnel, and a red oil bottle (though the oil bottle is not clearly visible, the red cap and label suggest it). In the background, three cars with open hoods are visible, matching the description. All core elements are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a red toolbox with various items on it, including two blue Peak antifreeze bottles, one white water bottle, one red funnel, and one red oil bottle.which the label bottle is not labeled labeled). it red color suggests label suggest it is Additionally the background, there cars with open hoods are visible, which the description. The elements elements of present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9adcc8c2fb7c992f72155b5223b32ec8408d87c1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce8c8b28c7958c545f7ab8702d74b1e7fc1a387270f05166f8eaee06dc31a91a
+size 72090
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..c1f7ed6c5990d428cd6827318216aca73e3a6adb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a gardening scene with a focus on two potted plants and a watering can. The watering can is green and has a white ribbon tied around it. The potted plants are placed on a wooden pallet, and the pots are white and blue. The plants are lush and green, indicating that they are well cared for. The watering can is positioned between the two pots, suggesting that it is being used to water the plants. The overall style of the video is simple and straightforward, focusing on the gardening activity without any additional embellishments. The video likely captures the process of watering the plants, highlighting the care and attention given to the plants."
+    ],
+    "video_ids": [
+        "0T0XntBbfZw_0_15to230"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two potted plants and a green watering can with a white ribbon.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two potted plants (one in a white container, one in a blue container) and a green watering can with a white ribbon tied around it. These elements are central to the scene and are accurately represented, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two potted plants inone green a blue pot and the in a blue container) and a green watering can with a white ribbon. around its. The elements match consistent to the description and match consistent represented. fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1c650b0eb9587c65336fe633cf80a39e467cd60a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:535822266e49ca5577a054fd530f9f0fb4650364c09e69cc69a3fd27d2d6f5d8
+size 88243
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..dbe4730e53307eeadac63e6481dc20c712007fed
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a body of water, possibly a river or a lake, surrounded by lush greenery. The water appears calm, reflecting the surrounding foliage and the sky above. The vegetation is dense, with tall grasses and reeds swaying gently in the breeze. The sunlight filters through the leaves, casting dappled shadows on the water's surface. The overall style of the video is naturalistic, with a focus on the tranquility of the scene. The camera seems to be positioned at a low angle, looking up at the towering grasses and reeds, which adds a sense of scale and grandeur to the image. The colors are vibrant, with the green of the foliage contrasting beautifully with the blue of the water. The video is likely to evoke a sense of peace and tranquility in the viewer."
+    ],
+    "video_ids": [
+        "qC4t-nF-0tw_68_0to110"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Water, lush greenery, tall grasses, reeds, sunlight, shadows? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by prominently featuring water in the background, lush greenery including tall grasses and reeds in the foreground, and sunlight casting natural shadows on the grasses and water. These elements align closely with the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting featuring water, the foreground, lush greenery, tall grasses and reeds, the foreground, and sunlight and shadows shadows on the wateres. water. The elements are with with the description provided.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d70c021c91d33da5470518a5915709de36231fde
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb4e8ad49868b2e1b28d8c92ae90079e3f7ef17e461ab2262ba79b62da67c741
+size 231828
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..8ebd8fe2d91edb305c9b16d368e6e417e4a1fc47
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a blue car, focusing on the driver's seat. The seat is black with blue stitching, and the car door is open. The car's interior is well-lit, and the driver's seat is positioned in the center of the frame. The car door is open, revealing the interior door panel and window controls. The car's interior is clean and well-maintained, with no visible damage or wear. The car appears to be a modern model, with a sleek design and high-quality materials. The video is a straightforward, unembellished depiction of the car's interior, with no additional context or narrative provided."
+    ],
+    "video_ids": [
+        "tlSBTuqywDM_26_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Driver's seat, car door, interior door panel, window controls? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the driver's seat with black and gray accents, the open car door, and part of the interior door panel. The window controls are not fully visible, but the presence of the door and seat aligns with the core description. The video does not contradict the specified objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows the driver's seat, a upholstery blue upholstery, the car car door, the the of the interior door panel. The window controls are partially visible visible, but the presence of the door handle the suggestss with the ' description of The video does not contradict the ' objects.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c107860f139c0631194a101fea5049085e0239f7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:804f25e2b2f50d7a29bd0deab2e992398ebe025981f77774e263bee5cb542188
+size 142798
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..22c08174097aefacea1fb33a292affedf423a154
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a pile of fresh green spinach leaves on a wooden surface. The leaves are vibrant and appear to be freshly washed, with some leaves overlapping each other. The focus is on the texture and color of the spinach, highlighting its freshness and natural beauty. The style of the video is simple and straightforward, with no additional elements or distractions. The lighting is bright and even, enhancing the natural colors of the spinach. The video does not contain any text or additional objects. The overall impression is one of simplicity and natural beauty, focusing on the freshness and appeal of the spinach leaves."
+    ],
+    "video_ids": [
+        "vsGMYPzpDqs_4_55to209"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A pile of fresh green spinach leaves? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by showing a clear, close-up view of a pile of fresh green spinach leaves. The leaves are vibrant, appear fresh, and are arranged naturally on a wooden surface, matching the description accurately. The only other element in the video is a brief clip of a man in a suit, which does not contradict the core description of the spinach leaves.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as depicting a pile and close-up view of a pile of fresh green spinach leaves. The leaves are vibrant and and to, and are piled in, a wooden surface, which the description of.\"\n There focus additional element present the video is the wooden glimpse of a wooden's a suit, but does not conflict the main description of the spinach leaves.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..434cde35b6598a96d433d3ccd4476d57767e0b50
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d75f8189c7f6eaa10b581036bdaf0e8cad4ffc5b9a7bf74678efc32e0c85be4f
+size 112773
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..f85017d052dabf27b14eac1e81e960a0ab0e49d8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a young man driving a red sports car with a black interior. He is wearing a blue t-shirt and glasses. The car is parked in a wooded area with trees and sunlight filtering through the leaves. The man is looking at the steering wheel and appears to be adjusting the settings on the car's dashboard. The car has a sleek design with a black roof and red accents. The man's expression is focused and he seems to be enjoying the experience of driving the car. The overall style of the video is realistic and it captures the excitement of driving a sports car in a natural setting."
+    ],
+    "video_ids": [
+        "e5yWk6MDjBc_51_0to192"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man, a red sports car with a black interior, a blue t-shirt, and glasses.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man wearing glasses and a blue t-shirt sitting inside a red sports car with a black interior. The car's interior features red stitching and a Ferrari logo, consistent with a high-performance sports car. The man is interacting with the steering wheel and dashboard, and the background shows trees, indicating an outdoor setting. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man driving a and a blue t-shirt, in a red sports car with a black interior. The car's design is a accents, a modern logo, which with the sports-end sports car. The setting is driving with the car wheel and gear, which the background shows a, indicating an outdoor setting. The elements elements of in present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3ae30837440ed0c4dca161505bd1e0b38abbb0a7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21bf216ae12e15f79c80e43dbecf8fb1d76cb3d97d54acdc3529b39de5f5eb87
+size 272460
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..6032aaa1fc087729be2dd1e1d30cbc210b72f09e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and a baseball cap, sitting in a dark room with a brick wall in the background. The man is wearing a gray shirt with a pattern of small white dots. The lighting in the room is dim, with a soft glow coming from the left side of the frame. The man appears to be in deep thought or contemplation, as he gazes off to the side. The overall style of the video is moody and introspective, with a focus on the man's facial expression and the texture of the brick wall in the background."
+    ],
+    "video_ids": [
+        "EjQ8BEr4cYs_0_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and baseball cap, wearing a gray shirt with small white dots.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and a baseball cap, wearing a gray shirt with small white dots, which matches the description. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and a baseball cap, wearing a gray shirt with small white dots. which matches the description provided The lighting appears lighting elements in not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..778551714dc1872c4f410ff9c7d0264628dee476
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1253-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c484c46791f5ea55eaa86953c627f9969faab53235db9a88f0f8e5c6748c9e7
+size 150715
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..93b261342bf72813c2e2f2df41ed4f67b55cdafa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video presents a stunning view of Earth from space, showcasing the planet's continents and oceans. The first frame captures the Earth from a distance, with the continents and oceans clearly visible. The second frame zooms in closer, revealing more details of the planet's surface. The third frame provides an even closer view, allowing for a more detailed look at the continents and oceans. The video is a beautiful representation of our planet, highlighting its vastness and the intricate details of its surface."
+    ],
+    "video_ids": [
+        "uh3hnKwAbTs_7_22to201"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Earth, continents, oceans? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows Earth from space, with visible continents (North and South America) and oceans surrounding them. The depiction is consistent with the requested 'Object(s)' condition, and no elements contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts two with space, with visible continents andsuch America South America, and oceans. them. The depiction of styl with the ' 'Object(s)' condition, as the elements contradict this description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2b5706f3c73e5ba793b18f38a0670cbcccab505d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8e1d700ecf5daa9f29dfbef3d543569064158cc9b325fb3fc79740020a4663df
+size 147028
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..203c0d9f8b773ce22b310894684ee22fea4469a1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the vibrant beauty of a sunflower field in bloom. The sunflowers, with their bright yellow petals and dark brown centers, are the main focus of the video. They are arranged in neat rows, creating a sense of order amidst the natural beauty. The sunflowers are in various stages of bloom, with some fully open and others still in bud form. The video is taken from a low angle, looking up at the sunflowers, which adds to the grandeur of the scene. The sun is shining brightly, casting a warm glow on the sunflowers and highlighting their vivid colors. The overall style of the video is naturalistic, capturing the beauty of the sunflower field in its natural state."
+    ],
+    "video_ids": [
+        "0zLomajgD0w_8_50to199"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Sunflowers (bright yellow petals, dark brown centers) in various stages of bloom.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows numerous sunflowers with bright yellow petals and dark brown centers, in various stages of bloom, including fully open flowers, buds, and some that appear to be wilting or past their prime. The description is accurately reflected in the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a sunflowers with bright yellow petals and dark brown centers, which various stages of bloom. which fully open flowers and partially, and partially partially appear to be wilting. dead their peak. The vibrant of largely represented in the visual content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a2ce4c772c0d0a76784855e710acad3c4729b1a7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68a0537a9d3af69fc55bca9d12d8ebc1aaf72825fd1d12cb9784e418ffbc79ae
+size 226822
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..8de10f402fbcc1096e14b15fb34c646f345fa0a5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a small dog, likely a Cavalier King Charles Spaniel, standing in the snow. The dog is wearing a red and black coat with a hood, which is designed to keep it warm in the cold weather. The dog's fur is a mix of brown and white, and it has long, floppy ears. The dog appears to be looking off to the side, possibly at something or someone out of frame. The background is a blur, but it seems to be a snowy landscape, suggesting that the video was taken during the winter season. The style of the video is a close-up shot of the dog, focusing on its face and upper body. The camera angle is slightly above the dog, giving a clear view of its coat and expression. The overall mood of the video is calm and serene, with the snowy setting adding a sense of tranquility."
+    ],
+    "video_ids": [
+        "N5kriwqfNMM_6_0to103"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Small dog (Cavalier King Charles Spaniel)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a small dog with the distinct physical characteristics of a Cavalier King Charles Spaniel, including its floppy ears, expressive eyes, and brown-and-white coat. The dog is appropriately sized and appears to be of the breed, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dog dog that the characteristic features characteristics of a Cavalier King Charles Spaniel, including its size ears, curly eyes, and the and-white coat. The dog is wearing dressed for the to be wearing the breed mentioned fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7ca5c712a35df12f7e5ea5391fae71fae2e5a02e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:58d5db09a0fab7a81bcdc8d008b03ebc5acc9f5745903e02ede78be91e228581
+size 159176
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..4674bfa60545e1a0ca7bc76c509218749ef01e4b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a man in a suit and tie, who appears to be a news anchor or reporter. The man is seated in front of a backdrop that includes a window with a view of a cityscape. The video is likely a news report discussing the marriage of Princess Eugenie, the youngest daughter of Sarah Ferguson and the Duke of York. The man is speaking, and his facial expression suggests that he is engaged in a conversation or delivering information. The style of the video is typical of a news broadcast, with a focus on delivering information in a clear and concise manner. The setting suggests that the video was filmed in a studio or newsroom."
+    ],
+    "video_ids": [
+        "tJvPcFLqtI4_7_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, seated in front of the backdrop.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a suit and tie, seated in front of a backdrop that includes studio elements like screens and windows, which aligns with the description. The man is clearly visible and the setting matches the specified condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed a suit and tie, seated in front of a backdrop that appears a elements like a and a, which ares with the description of The presence is positioned the and appears setting is the typical conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c0e372dd8ae8743b70395a4fdec428666c9f56f8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30b6d234b2fcc23d39725fb62cde5ad4280510ef0c90110e50e665a946616348
+size 141249
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..ba6eebe7a5e83dd69087f2aa96712d80542bc482
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment in a car, where an older man and a younger man are engaged in a conversation. The older man, wearing a white shirt and glasses, is driving the car, while the younger man, dressed in a blue and white plaid shirt, is seated in the passenger seat. The car is moving along a road, with trees visible in the background. The interaction between the two men suggests a friendly or familial relationship, as they seem to be sharing a moment of connection. The overall style of the video is candid and unposed, capturing a slice of everyday life."
+    ],
+    "video_ids": [
+        "Lr_aZ20W4k4_105_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An older man driving and a younger man seated in the passenger seat.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an older man in the driver's seat holding the steering wheel and a younger man seated in the passenger seat. Both individuals are clearly visible and positioned as described, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an older man driving the driver's seat of the steering wheel, a younger man seated in the passenger seat. Both are are wearing visible and positioned as described, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..920305dcd7455afa6b8299834aee8e5969aec2c2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:78b3d493f705a9fe8ed8581ae39b544e111d91f9094bee3f18dd46a1bacf3b60
+size 188700
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..da6de2fa29fd22bf3865fcf436019defad90003b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man sitting in the back seat of a car, speaking into a microphone. The car's interior is visible, with leather seats and a door on the right side. The man is wearing a blue shirt and appears to be engaged in a conversation or recording a podcast. The style of the video is casual and informal, with a focus on the man and his surroundings. The lighting is natural, suggesting that the video was taken during the day. The man's expression is serious, indicating that he is concentrating on his speech or discussion. The overall impression is that of a personal or professional video, possibly for a podcast or interview."
+    ],
+    "video_ids": [
+        "23i__Z-Hm9c_61_65to192"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man, microphone? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man speaking, and a microphone is visible clipped to his shirt. These elements match the 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man sitting into and there microphone is positioned in to a shirt. The elements fulfill the 'Object(s)' condition specified any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1c3172e03c26c537ff6e990eadef021550d2cc43
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:627b2a5a766003c0b0b3a9bdde0e9f9643a902d0f953e238a1ce080f7cdff74d
+size 132313
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..b7b2202f283358ae6ff388a3133913eddd69eb1e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen interacting with a black sports car. The car is parked on a road with a grassy field in the background. The man, dressed in a white shirt and jeans, is first seen walking towards the car. He then leans over the hood of the car, inspecting it closely. Finally, he reaches out to touch the car, perhaps admiring its sleek design or checking for any imperfections. The video captures the man's interest in the car, suggesting a narrative of admiration or ownership. The setting, with its open field and clear sky, adds to the sense of freedom and adventure often associated with sports cars."
+    ],
+    "video_ids": [
+        "O6KnDYgQQgo_69_0to122"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black sports car and a man dressed in a white shirt and jeans.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a black sports car and a man dressed in a white shirt and jeans, which matches the described 'Object(s)' condition. The man interacts with the car by opening the hood, and the background elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a black sports car, a man dressed in a white shirt and jeans. which align the description 'Object(s)' condition. The man is with the car by opening the hood, which the setting includes, not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..86815ea5ffe2d81709b05b2d3d03ab77f1a440a9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef85167b0946f04fa8c3defff649232271824a17566ba3728b3e4e276e34ee91
+size 233985
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..9766a83c547d2f2c687edfe7db93a58075539bd5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in front of a brick wall with a television screen behind him. He is wearing glasses and a gray hoodie with a white logo on the front. The man appears to be speaking or presenting, as he is gesturing with his hands. The television screen is turned on, displaying a colorful interface, possibly a computer or gaming screen. To the right of the man, there is a potted plant with green leaves, and a small, round, white lamp with a yellow light. The overall style of the video seems to be a casual, informal setting, possibly for a video blog or a casual presentation."
+    ],
+    "video_ids": [
+        "0fX6LhvmcB0_35_64to222"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man, brick wall, television screen, potted plant, small white lamp? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a grey hoodie, a brick wall in the background, a television screen displaying content, a potted plant, and a small white lamp. All specified objects are present and accurately represented in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a man standing a gray hoodie standing standing brick wall in the background, a television screen, a, a potted plant, and a small white lamp. These the objects are present and match depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f20436120f5ce67d0a73182e4c47d88519ca2f8d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1254-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e9260f9185038c7f6b9acef712eecc2ffb91333a5e3d6e6ee12e6a80b166ae60
+size 171505
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..5583473323e8c5580d1566ca7e48c99c7bf6ad8f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment from a hockey game, featuring players from the Winnipeg Jets. The main focus is on a player wearing a blue jersey with a red maple leaf logo, who is holding a hockey stick. He is surrounded by his teammates, who are also wearing blue jerseys with the same logo. The players are standing on the ice, and the background shows the stands filled with spectators. The style of the video is a standard sports broadcast, with a focus on the action on the ice. The camera angle is from the side of the ice, providing a clear view of the players and their movements. The lighting is bright, highlighting the players and the ice. The overall atmosphere is energetic and competitive, as is typical for a hockey game."
+    ],
+    "video_ids": [
+        "2TZLpbLWEDs_18_149to270"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a blue jersey with a red maple leaf logo, holding a hockey stick; his teammates in similar attire, also on the ice.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows players in blue jerseys with a red maple leaf logo, consistent with the Winnipeg Jets' uniform. The central player is wearing a jersey with the maple leaf and is surrounded by teammates in similar attire. Although the hockey stick is not clearly visible, the context of the scene (hockey players on the ice) and the presence of hockey gear (gloves, helmets) strongly imply the players are holding or have access to hockey sticks. The description is largely accurate.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a wearing blue jerseys with a red maple leaf logo, holding with the description Jets' uniform. They players player is holding a helmet with the team leaf logo holding holding by teammates in similar attire. They the video stick is not clearly visible in the players and the players stronglyplayersockey players on the ice) and the presence of other helmets (helloves and helmets) supports suggest that player are holding hockey are recently to hockey sticks. The video is largely fulfilled,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f27aaefc8a846e70f7f450f471187c72e59fe9a3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee5a4ad4cb86d3265e96995be43295f4bb55c45e3a6aa8e4d60ad76f6971cf02
+size 323142
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..72bcac37a3dd7a6dab8e35a45d8fc3c09231db0d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the beauty of a tree blooming with white flowers in College Park, Maryland. The flowers, with their delicate petals and vibrant green centers, are the main focus of the video. The tree is set against a backdrop of a brick building, adding a touch of urban charm to the scene. The video is shot in a close-up style, allowing the viewer to appreciate the intricate details of the flowers. The colors in the video are predominantly white, green, and red, creating a harmonious and pleasing visual palette. The overall style of the video is simple yet elegant, focusing on the natural beauty of the flowers and their surroundings."
+    ],
+    "video_ids": [
+        "GWiTBJ7vdpA_2_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A tree blooming with white flowers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows close-up shots of white flowers with pinkish edges, which are characteristic of dogwood blossoms. These flowers are attached to branches, indicating they are part of a tree. The background includes green foliage and branches, consistent with a blooming tree. The text overlay mentions 'University of Maryland College Park, Maryland,' which is a plausible location for such a scene. The core description of a tree blooming with white flowers is accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a-up shots of white flowers with greenish centers, which are likely of awood blossoms. The flowers are blo to a, and a are part of a tree. The presence includes a leaves and a, which with a treeoming tree. The presence ' mentioning 'dog of Washington,' Park,' MD, which is not location location for such a scene, The video description of a tree blooming with white flowers is fulfilled represented in}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3d3a900121fe695720afd75dbd694837c5a6f95b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1c84012419bbc02e14a7e828a639ce49d4ae367b1a0b5b991d48f942451d1eac
+size 216303
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..0f52d5b8a3672b73db03efce58f6263815bf5187
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a car, focusing on the steering wheel and dashboard. The steering wheel is black with a silver emblem in the center, and the dashboard is sleek and modern, featuring a touch screen display. The car's interior is well-lit, with the sunlight streaming in through the windows, highlighting the car's luxurious materials and finishes. The car appears to be parked in a lot, with a building visible in the background. The video is shot in a realistic style, with a focus on the car's interior and the surrounding environment."
+    ],
+    "video_ids": [
+        "WF2n_chkGPE_48_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, dashboard, car's interior, windows, building in the background? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the steering wheel with the Buick logo, the dashboard with gauges, the car's interior including the center console and door panels, windows (visible on the sides), and a building in the background through the windshield. All specified elements are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a steering wheel, the Genesisick logo, the dashboard with variousuges and the car's interior with the seats console and seats panels, the,part in the sides), and a building in the background. the windows. The these elements are present and match depicted,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c3f17ab941536f336a459586b332dc3e8d2e7392
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5f93b3852c96ba09bd6db10a1bba3eedcfa6b27d7827e303fd08ccf2b99fb4f
+size 226227
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..70c8fc20f544e952ef873ef96614a747063e821c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a colorful motorcycle helmet with a vibrant design featuring flames and a dragon. The helmet is placed on a white surface, possibly a table or a countertop. In the background, there is a person, likely a man, who appears to be gesturing or explaining something. The style of the video seems to be a casual, informal presentation, possibly for a product review or a tutorial. The focus is on the helmet, with the person in the background providing context or commentary. The lighting in the room is bright, and the overall atmosphere is casual and relaxed."
+    ],
+    "video_ids": [
+        "0hwn-sWtsk4_2_106to244"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A colorful motorcycle helmet with a vibrant design featuring flames and a dragon, a person (likely a man).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a colorful motorcycle helmet with a vibrant design that includes flames and a dragon, matching the description. A person (likely a man) is also visible in the background, gesturing towards the helmet, which satisfies the 'person' condition. The background elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a colorful motorcycle helmet with a vibrant design that includes flames and a dragon. which the description. Additionally person,likely a man) is also visible in the background, addinguring towards the helmet, which ful the 'Object' condition. The presence is do not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cec357d511478a4699c6ffdf6ed624cb3c9e8174
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2febfdafe23297b7ff7599d480b147c69305e4677853e8bc8c1e57549e9f9ba8
+size 219209
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..d798cd33e4e56a5f26b4baa455b555f9c49b4e88
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a playful and imaginative mashup of animated characters from different movies and universes. It features a gold robot standing confidently in the center of a room filled with animated princesses. The princesses are dressed in colorful gowns and are engaged in various activities, such as dancing, playing, and conversing. The room is decorated with elegant furniture and a chandelier, creating a festive atmosphere. The robot's presence adds a touch of humor and surprise to the scene, as it contrasts with the traditional fairy-tale setting. The video is likely designed to entertain and amuse viewers by combining elements from different genres and franchises."
+    ],
+    "video_ids": [
+        "lJ-hbJDWazY_14_24to160"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A gold robot and several animated princesses engaged in various activities.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a prominent gold robot (C-3PO) and multiple animated princesses from Disney films, all engaged in various activities such as dancing, standing, and interacting. The scene is set in a luxurious room, and the characters are clearly visible and consistent with the described condition. No elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a gold gold robot standinglikelyheet3PO) standing several animated princesses in Disney movies. engaging of in what activities such as dancing and which, and interacting with The setting is set in a grand room with which the characters are dressed animated, distinct with the description elements. The additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..56e42f258b3b1d78264fae0462c40941f9ddd7a3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:285d2486c4ccdbd42893ba19d7c5b34591af3a177256788ced61f798797e06f9
+size 300023
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..c04973e80b437e9bb7eb9f4200736bd4de49d623
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse of a car's interior, showcasing the dashboard and steering wheel. The car's interior is luxurious, with a wooden dashboard and a large touch screen display. The steering wheel is black and features a logo in the center. The car is driving through a forested area, with trees visible through the windshield. The time-lapse effect creates a sense of motion and progression, as the car moves through the landscape. The overall style of the video is sleek and modern, with a focus on the car's interior and the natural scenery outside."
+    ],
+    "video_ids": [
+        "Bu0UhBrzd8k_13_0to119"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A car with a luxurious interior featuring a wooden dashboard, a large touch screen display, and a black steering wheel with a logo in the center.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows the interior of a luxury car with a wooden dashboard (orange-brown wood trim), a large central touch screen display showing a map, and a black steering wheel featuring a central logo (Rolls-Royce emblem). These elements align with the described 'Object(s)' condition, even though the video does not show a person or any other additional elements that contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a interior of a car car with a black dashboard,though accentsrown tones visible visible a large touch touch screen display, a navigation, and a black steering wheel with a logo logo.likelys-Royce).). The elements align with the description 'lux(s)' condition, making though the video is not show the full driving the additional elements elements.\"\n might the core.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8c0d27c970e3cdd97650cfff6fe6de63a3dde61f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:670a8bdaee842640f0ef85eb97a36418a789a4d6440c2ce1cc9c2bfc8ca2b16a
+size 235840
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff94d07150fc5bd432fc6f9d6e833cad9af1003b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing on a stage, speaking into a microphone. He is dressed in a dark suit and a white shirt, with his sleeves rolled up. The man appears to be in the middle of a speech or presentation, as he holds a piece of paper in his hand. The background is dark, which puts the focus on the man and his actions. The style of the video is straightforward and professional, with no additional elements or distractions. The man's attire and the microphone suggest that this is a formal event or gathering. The video captures the man's speech in progress, providing a glimpse into the event and the speaker's message."
+    ],
+    "video_ids": [
+        "ap1N1yFTsHs_3_0to177"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man, microphone, piece of paper? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a microphone and holding a piece of paper, matching the specified objects. The background is plain black, which does not conflict with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man standing a suit, holding a piece of paper, which the ' objects. The man and minimal,, which does not contradict with the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ed2475ce77d11e998ae76f56ad5e0e8dc9741cd0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41b5c06b501f3971319056df78908720cac2e5b816579302f796a366f203347a
+size 114274
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..3b2a1c7e332b6075734ad5dfc9b80224fa85b4ad
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person standing next to an open car trunk. The car is parked on the side of a road, and the person is wearing a gray sweatshirt and blue jeans. The trunk is empty, and the person appears to be looking inside it. The background of the video shows a hillside with trees and a clear sky. The style of the video is casual and seems to be taken with a smartphone or a similar device. The focus is on the person and the car trunk, with the background serving as a simple backdrop."
+    ],
+    "video_ids": [
+        "OWx65twUmyg_25_20to234"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, an open car trunk? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person interacting with an open car trunk. The person is seen bending over and adjusting the trunk's interior, and the trunk is fully open, revealing its spacious interior. There are no elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a person standing with an open car trunk. The person is standing standing down and reaching something trunk, contents, which the trunk is open open, fulfilling the contentsness. The are no additional in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d27b5803db0d062cd92ada0c9a83558b839c644b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1255-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:388652ad0d9d30a47276c46dee4675745101106e86eef305b3f0ebf4439b85e1
+size 197919
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..429bcb17af3ba1f451cd60253a03be7a242152f6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene inside what appears to be an indoor zoo or aquarium enclosure. A polar bear is the central focus, positioned near the glass barrier that separates it from the viewers. The bear is partially submerged in clear blue water, with its head and upper body visible above the surface. Its fur is thick and white, contrasting sharply with the blue water. The bear's movements are slow and deliberate; it seems to be gazing out at the visitors or perhaps observing something beyond the glass.  In the background, the enclosure is designed to mimic a natural environment, featuring rocky structures and greenery that provide a sense of depth and space. The rocks are light gray and textured, adding to the authenticity of the setting. The greenery includes trees and bushes, which are lush and vibrant, suggesting a well-maintained habitat. The sky outside is overcast, with a soft, diffused light that filters through the clouds, creating a calm and tranquil atmosphere.  As the video progresses, the"
+    ],
+    "video_ids": [
+        "6200ecee770a19f26be74b05119cdac011b0ba750b0df6a25c63ca4063e923ca"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A polar bear? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a polar bear in an enclosure with water, interacting with a glass wall. The bear's white fur, size, and behavior are consistent with a polar bear, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a polar bear in a aquatic that water, which with its pool barrier. The bear is white fur and shape, and posture are consistent with the polar bear. fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..852f374fc0600f41cc1a9a55578df8f79fc6940d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:61e294324d212acd440b8e2d303fc4cd88caf8f77926e3565d003ab993240b6a
+size 250689
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..8ca32122a60866a79b069fe06079d92b3e1bd636
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the beauty of a rose garden in bloom. The roses, in shades of pink and white, are the main focus of the video. They are surrounded by lush green leaves, adding a vibrant contrast to the soft colors of the flowers. The roses are in full bloom, their petals unfurling to reveal the intricate details of their structure. The video is taken from a low angle, looking up at the roses, which adds a sense of grandeur to the scene. The roses are in the foreground, with the green leaves filling the background, creating a sense of depth and perspective. The video is a celebration of nature's beauty, capturing the fleeting beauty of a rose garden in bloom."
+    ],
+    "video_ids": [
+        "tuwkgsQzvnI_9_0to159"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Roses in pink and white, green leaves? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows flowers that resemble roses with pink and white petals, surrounded by green leaves, matching the description. Although the flowers may not be true roses (they appear to be begonias), they are visually consistent with the requested object types.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a that are roses in a and white petals, and by green leaves. which the description provided The the video are not be real roses,as could to be aonias or the still still similar with the given colors(s.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3bd35c757405f383f02cef4f9e93378b7082bb23
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c23d874c4611d179f2968c4778715187b0cd09fab5fbc220ab559ffa47722de4
+size 152729
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..eb1e4dbb955f14bcb2f4ac76f49209a53925d464
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a kitchen, preparing a meal. He is wearing a blue and white striped shirt and has tattoos on his arms. The kitchen is well-equipped with a stainless steel oven, microwave, and sink. There are various kitchen utensils and ingredients visible, including a cutting board, knives, spoons, and bowls. The man is seen washing a blender under the kitchen sink, and there are bottles of oil and vinegar on the counter. The style of the video is a casual, home-cooking tutorial, with a focus on the man's actions and the kitchen environment."
+    ],
+    "video_ids": [
+        "9g7W82cGRHI_9_0to162"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue and white striped shirt with tattoos, a cutting board, knives, spoons, bowls, a blender, bottles of oil and vinegar.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue and white striped shirt with visible tattoos on his arms. He is in a kitchen with a cutting board, knives, spoons, bowls, a blender, and bottles of oil and vinegar. These elements are all present and consistent with the description, even though the man is pouring something into a pot and then filling a glass with water, which does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a blue and white striped shirt with visible tattoos, his arm. He is in a kitchen setting a cutting board, knives, andons, bowls, and blender, and bottles of oil and vinegar. The elements match consistent present and match with the description provided indicating though the blender is not from from a pot instead not into a blender, water, which is not directly the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0e6f7057fffd245271ccd99a511a6684bd0247f7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:062d6578b270f7364ce18d1e69bdbfb7f2d0a9baff7d6d4e3b78e74c300b8c4e
+size 187379
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..7dc7eb1b58c849df476461733e8318f2967a8cbd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a vibrant underwater scene teeming with life. A school of colorful fish, predominantly green and orange with hints of red on their fins, swims gracefully through the water. These fish are interspersed with other species, including some darker-colored fish that contrast sharply against the brighter hues. The seabed is adorned with various marine plants and corals, adding texture and depth to the environment. The camera pans smoothly across the scene, revealing the dynamic movement of the fish as they navigate around the coral formations and aquatic vegetation. The lighting is natural, filtering through the water, which enhances the vivid colors of the fish and the surrounding marine life. The overall atmosphere is serene yet lively, showcasing the beauty and diversity of the underwater world."
+    ],
+    "video_ids": [
+        "f68d89540b741fefbca17c358706a6fc659b0186f8ce750a2d2adf5be02ac909"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: School of colorful fish (predominantly green and orange with red fins) and darker-colored fish.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a school of fish that are predominantly green and orange with red fins, which matches the description. Additionally, there are darker-colored fish visible, such as blue and black ones, which also align with the condition. The presence of coral and other underwater elements does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a school of colorful with are predominantly green and orange with red fins, which align the '. Additionally, there are darker-colored fish present in fulfilling as black and black,, which also align with the description. The presence of these and other marine elements does not contradict the core description but}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..763522fbe9ee65ef873649295801a22ab78bdc27
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f427cc6f5cfdcb02781a188a5910624608126e427008cbf41e29aa25408d3e5a
+size 319454
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..bf07e738485bef9cea9ee464865b4ec31a1cdf88
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a garage, wearing sunglasses and a gray shirt. He is holding a shoe in his hand, which is shown in a close-up shot. The garage is filled with various items, including a red fire extinguisher and a yellow toolbox. The man appears to be in the process of fixing or inspecting the shoe. The overall style of the video is casual and informative, with a focus on the man's actions and the objects in the garage."
+    ],
+    "video_ids": [
+        "MUC8bBtuCCU_18_121to332"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a shoe, a red fire extinguisher, a yellow toolbox? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man wearing sunglasses and a gray shirt, a shoe visible in an inset image, a red fire extinguisher (with a Texas license plate) in the background, and a yellow toolbox (or similar tool container) also visible in the background. All specified objects are present in the video, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man holding sunglasses, a gray shirt, holding black in in his open,, a red fire extinguisher,though a visible flag plate sticker on the background, and a yellow toolbox.with similar container storage) also in in the background. The elements elements are present and the scene, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..faef9264929c22ae5e8f97d8d16053e02a8238b5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a9616ed50adb8e6a31ba74dbba0cce4f4a5df9e41046ef2888a67e40361f4de
+size 188715
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..b9008c69d3355255ce62fcaef4811a293d134d93
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a woman in a white vest and black shirt, standing in a room with a group of people. She is speaking on a phone, her mouth open and her eyes focused on something off-camera. The room is filled with people, some of whom are seated at a table, while others stand around. The woman appears to be in the middle of a conversation, her expression serious. The room is well-lit, with natural light coming in from the windows. The people in the room are dressed in casual clothing, suggesting a relaxed atmosphere. The woman's white vest stands out against her black shirt, drawing attention to her as the main subject of the video. The room is filled with activity, but the woman on the phone is the focal point of the scene."
+    ],
+    "video_ids": [
+        "HH1M741TnJ0_46_0to166"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a white vest and black shirt, a group of people in the room.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a woman wearing a white vest over a black shirt, which matches the core description. Additionally, there are other people visible in the background within the room, fulfilling the 'group of people in the room' condition. The presence of other elements, such as tables, chairs, and decorations, does not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a woman wearing a white vest and a black shirt, which matches the description description. Additionally, there is other people in in the background, the same, fulfilling the 'group of people' the room' condition. The presence of these individuals in such as the and chairs, and windows, does not contradict the core and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d15d5bcd06b89ba1b83b8d1047780c953fda893d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:590be5bfa1aeda37443c70220287720c9f9f2e2b578845a23e42fa9431915ddd
+size 141820
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..e869186ae8c1dd9ac9330761e2c4cb858e38b7cf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up shot of the interior of a car, focusing on the steering wheel and dashboard. The steering wheel is prominently displayed in the center of the frame, with the Audi logo clearly visible. The dashboard features a digital display screen, which is illuminated with various icons and information. The car's interior is sleek and modern, with a black and silver color scheme. The video is likely a promotional or informational video for the Audi brand, showcasing the design and features of the car's interior. The style of the video is professional and polished, with a focus on the car's design and technology."
+    ],
+    "video_ids": [
+        "i--fke8G948_10_0to170"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel (with Audi logo), digital display screen on the dashboard, icons and information on the display screen.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a steering wheel with the Audi logo prominently displayed in the center. A digital display screen is visible on the dashboard, and various icons and information are displayed on the screen, matching the description. The presence of other elements like the gear shift and air vents does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a steering wheel with the Audi logo, displayed in the center. The digital display screen is visible on the dashboard, and it icons and information are present on the screen, which the description provided The presence of the elements like the gear shift and part vents does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d322d32f095c3cfc2f62c12635d5458693741cf3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:423856153afbd85de3f5d718fcd6914914bc22950dac3b0e9eba404c9d7eed42
+size 135432
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..e848a85e43b27e94191c478e10b0989717235910
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, three people are gathered in a kitchen, engaging in a cooking activity. The kitchen is well-equipped with wooden cabinets, a sink, a stove, and a microwave. The counter is adorned with various cooking utensils and ingredients, including a plate of food, a bowl, and a cup. The people are standing around the counter, with one person actively cooking on the stove, while the other two are observing and possibly discussing the cooking process. The overall atmosphere of the video suggests a friendly and collaborative cooking session."
+    ],
+    "video_ids": [
+        "QAIDeOcqbGI_48_0to197"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three people, a plate of food, a bowl, a cup, a stove, a sink, wooden cabinets, a microwave, and cooking utensils.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three people standing in a kitchen setting. On the counter, there is a plate of food (pancakes), a bowl (with whipped cream), and a cup (with syrup). A stove is visible with pans on it, and a sink is in the background. Wooden cabinets, a microwave, and cooking utensils are also present, matching the described elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts three people in in a kitchen.. There the counter, there is a plate of food,whichcakes and a bowl,possibly a cream), a a cup (possibly a). The stove is visible in a on it, and the sink is partially the background. The cabinets are a microwave, and various utensils ( also present. fulfilling the description elements.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3a61d8842362d2ceb47fbbb9eac3bf00eb87b0b7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1256-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17d6b717c0aa9b7523114f1c003b7b5cb1f6dfd744e3a54d01c23e405f1bbd9c
+size 111390
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..2ef5669d7a3d05f5f4c7a108b787382028353b88
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man enjoying a meal at a restaurant. He is seated at a table, wearing a black t-shirt, and appears to be in the middle of a conversation. The man is holding a sandwich in his hands, taking a bite, and savoring the taste. The sandwich is generously filled with ingredients, suggesting it might be a sub or a burger. The restaurant has a casual atmosphere, with other patrons visible in the background. The lighting is warm and inviting, creating a comfortable dining environment. The man's enjoyment of his meal and the relaxed setting suggest a pleasant dining experience."
+    ],
+    "video_ids": [
+        "RYHs4tV3vcY_23_62to242"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a table, a sandwich? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man eating a sandwich at a table in what appears to be a casual dining setting. The core objects \u2014 a man, a table, and a sandwich \u2014 are all present and accurately depicted. Additional background elements (other people, ceiling fans, etc.) do not contradict the description and are consistent with a public dining environment.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man sitting a sandwich at a table. a appears to be a restaurant dining setting. The presence elements ( a man, a table, and a sandwich \u2014 are all present and match depicted. The elements elements likelike din, restaurant lights, and.) do not contradict the main.\"\n are typical with a typical dining environment.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..aeb2299f4b17e654189d2775aa6ed491e6efedc5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52048c3e72574b77052ac292ce9647c3d525f6dd8be5b65b3d464103e674d209
+size 180762
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..7e0dcd448156393307539fab24f1f9d2164038ef
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young woman standing in front of a gold-colored train car. She is wearing a blue and white polka dot dress and has a brown purse slung over her shoulder. The train car is labeled \"Sweet Train\" and has a heart symbol on it. The woman is smiling and appears to be posing for the camera. The train car is parked on a platform, and there are other people visible in the background. The overall style of the video is casual and seems to capture a moment of travel or adventure."
+    ],
+    "video_ids": [
+        "L5Maal7BZAw_83_17to172"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman, a blue and white polka dot dress, a brown purse, a gold-colored train car, and other people in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman wearing a blue and white patterned dress (resembling polka dots) with a brown strap visible, likely from a purse. Behind her is a gold-colored train car with 'JRKYUSHU SWEET TRAIN' written on it. In the background, there are other people and structures consistent with a train station environment, fulfilling the 'other people in the background' condition. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a young woman wearing a blue and white poled dress,notsembling polka dots) and a brown purse purse on standing from a purse. She her is a gold-colored train car with aS'OTOU'AKET'' written on it, There the background, there are other people, green that with a train station setting. fulfilling the 'other people in the background' condition. The elements elements of in present and contradiction.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..061ecf88b2a35a7fdc18ab969e5a0e54cc6dbb03
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6ad238888c804baaa9517cace076460ff0c59c551e229f2627fba5c2c69b4d6d
+size 261246
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..3b408501a317101dd9242c0c09ab0f1401b7ac8a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a white sports car driving down a road near a body of water. The car is sleek and modern, with a black spoiler on the back and black rims on the wheels. The car is moving at a moderate speed, and the driver is focused on the road ahead. The road is lined with a fence on the right side, and there are trees and bushes on the left side. The water is calm and blue, reflecting the clear sky above. The sun is shining brightly, casting a warm glow on the scene. The overall style of the video is dynamic and energetic, capturing the thrill of driving a sports car on a beautiful day."
+    ],
+    "video_ids": [
+        "1z2pPRqCMK4_7_0to110"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a white sports car, specifically a modified Mercedes-Benz C63 AMG, parked on a paved surface. The car's design, including its rear spoiler, custom wheels, and sporty body kit, aligns with the description of a white sports car. Additional elements like the fence and background scenery do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a white sports car, which a white sports-Benz AM63 AMG, driving on a road area. The car's design, including its sleek spoiler and exhaust wheels, and overally appearance kit, aligns with the description of a sports sports car. The elements such the ocean, the scenery do not contradict the core description but}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b8409dcd3845b0073a042d2a8b7024b67fe6aa7c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:380eed30324e2f787686f67d3c0d66e71b5ee590a33a1fe0bcf3a7561fab48ef
+size 303628
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..3af3f7734c733357458e2332576e31e05fc531a8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a collection of automotive engines on display at a trade show. The engines are meticulously arranged on a black countertop, each one encased in a red cushion for protection. The engines vary in size and design, reflecting the diversity of automotive technology. The background features a white wall adorned with informational posters, providing context and details about the engines. The overall style of the video is informative and promotional, aimed at showcasing the latest advancements in automotive engineering."
+    ],
+    "video_ids": [
+        "BrwSGdOHvYI_16_16to185"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Automotive engines encased in red cushions.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows several automotive alternators (a type of engine component) displayed on red cushions, which matches the description. The objects are clearly visible and presented in a way that aligns with the stated condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple automotive enginesators encnot type of engine component) enc on red cushions. which align the description of The presence are indeed enc and are in a manner that aligns with the given condition.}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9916fcf4edacbfa87f4658b6c3e8690061cdc852
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:66cf92baff04407890b51fdd2f969630a6f672c0bd67f91183f8cbf5448c7c98
+size 172409
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..252f0210570d04bf935bce4bca7b79cfe10ec7db
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a bartender in action, crafting a cocktail in a bar setting. The bartender is seen in the background, pouring a drink into a glass. The glass is placed on a bar counter, and it's filled with a brown liquid, possibly whiskey or a similar spirit. The bartender adds a twist of orange peel to the drink, adding a touch of citrus to the cocktail. The bar counter is made of wood, and there are shelves in the background filled with various bottles, suggesting a well-stocked bar. The lighting in the bar is warm and inviting, creating a cozy atmosphere. The video is likely shot in a real bar, capturing the bartender's skill and the ambiance of the establishment."
+    ],
+    "video_ids": [
+        "QGZF-thxUWo_1_22to167"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bartender, glass, orange peel, bottles on the shelves? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bartender placing an orange peel on a cocktail glass, with bottles visible on shelves in the background. All specified objects are present and clearly depicted in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bartender pouring an orange peel into a glass in, which bottles visible on the in the background. The the elements are present and the depicted, the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..59e4b05f9c486afa45b5a1c2a7f6393a9fab678c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0ce08cae0ab48acd94868942bc49ea4d163981d59d38a3c460dc9bd8d96c25a9
+size 153259
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..6cb3633a0c06e3447a3b785031388cb4ffcb85e8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a colorful motorcycle helmet with a vibrant design featuring flames and a dragon. The helmet is placed on a white surface, possibly a table or a countertop. In the background, there is a person, likely a man, who appears to be gesturing or explaining something. The style of the video seems to be a casual, informal presentation, possibly for a product review or a tutorial. The focus is on the helmet, with the person in the background providing context or commentary. The lighting in the room is bright, and the overall atmosphere is casual and relaxed."
+    ],
+    "video_ids": [
+        "0hwn-sWtsk4_2_106to244"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A colorful motorcycle helmet with a vibrant design featuring flames and a dragon, a person (likely a man).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a colorful motorcycle helmet with a vibrant design that includes flames and a dragon, matching the description. A person (likely a man) is also visible in the background, gesturing towards the helmet, which satisfies the 'person' condition. The background elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a colorful motorcycle helmet with a vibrant design that includes flames and a dragon motif which the description. Additionally person,likely a man) is also visible in the background, whichuring towards the helmet, which ful the 'person' condition. The presence elements, not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d5b3f2aa60510d8e3fd28bc7ba54eb2bfa53d5b3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:339f8b19a0b26abf5f49e652273b755d3eea2e7a8a89e358ea954b17a3235cb9
+size 168245
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..cfc02e9e0bb2b76df96c95853778154a17ed5cdd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a bustling city street filled with various vehicles, including cars and trucks, all moving in the same direction. The street is lined with trees and traffic lights, adding to the urban atmosphere. The vehicles are of different colors and models, suggesting a diverse group of drivers. The street appears to be busy, with traffic moving smoothly under the guidance of the traffic lights. The overall style of the video is realistic, capturing a typical day in a busy city."
+    ],
+    "video_ids": [
+        "cpmRL1cZCHY_29_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Various vehicles including cars and trucks? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple vehicles including cars and trucks moving along a city street. Various models and colors of cars are visible, including sedans, SUVs, and a convertible, which aligns with the 'Various vehicles including cars and trucks' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a vehicles, cars, trucks, through a busy street. The types of colors of cars are visible, and sedans, hatchs, and a truck, which aligns with the descriptionVarious vehicles including cars and trucks' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d62040eda8c638eb45fed7886dbab6407565b8a3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:319ea8581d1057f59a51055984ebf9ae84fa9622b5acb40b640348a19f38a1e7
+size 183895
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..6b14010764a0a9b54e2df371d539f6abb1518cb9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a dynamic and intense comic book style animation featuring a robotic character. The robot, with its blue and silver armor, is the central figure in the video. It has a menacing appearance, with red eyes that glow menacingly. The robot is shown in three different frames, each capturing a different action. In the first frame, the robot is shown in a close-up, its red eyes glowing brightly. In the second frame, the robot is shown in a mid-action pose, its arm extended as if it's about to strike. In the third frame, the robot is shown in a full-body shot, its arm raised high, ready to unleash its power. The background of the video is a fiery explosion, adding to the intensity of the scene. The overall style of the video is reminiscent of a comic book, with bold lines and vivid colors. The robot's design is intricate, with details that suggest advanced technology. The animation is fast-paced and action-packed, capturing the viewer's attention from the very beginning."
+    ],
+    "video_ids": [
+        "fw-lNSCNXBY_15_0to123"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A robotic character with blue and silver armor, red glowing eyes, and advanced technology details.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a robotic character with blue and silver armor, red glowing eyes, and detailed advanced technology features, matching the description. The character's design includes mechanical elements, glowing red eyes, and a futuristic aesthetic consistent with advanced technology. The presence of a fiery background does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a robotic character with blue and silver armor, which glowing eyes, and advanced advanced technology features. such the description provided The character's design, a elements and glowing eyes eyes, and a futuristic appearance, with the technology. The background of a fiery background does not contradict the description description but}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..12a3907a655102d65002e49f0075baf1c4636529
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1257-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:478fe07597999bac1b9e048fdf8e015abf6a0ae252cf82fd5f17ffb849696138
+size 445081
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..b37d613fb256425d9a97d2bafce03a281c5fe309
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a scientific experiment involving a glass apparatus with various tubes and flasks. The apparatus is connected to a pump and a pressure gauge, suggesting that it is used for measuring and controlling pressure. The tubes and flasks are filled with a white substance, which is likely a chemical or a biological sample. The experiment appears to be in progress, as the pressure gauge is showing a reading. The style of the video is educational and informative, aimed at demonstrating the use of scientific equipment and techniques. The focus is on the apparatus and the experiment, with no additional context or background provided. The video is likely intended for an audience interested in science or chemistry."
+    ],
+    "video_ids": [
+        "8GVSuKkuLzY_12_368to549"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Glass apparatus, tubes, flasks, white substance, pump, pressure gauge? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a glass apparatus with tubes and flasks, a white substance inside the flask, and a pump system with tubing and red caps. Although a pressure gauge is not visibly present, the core elements described are largely present and consistent with the setup shown.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a glass apparatus with tubes and flasks, containing white substance ( a fl, and a pump with with a. a markings. The the pressure gauge is not visible present, the setup elements of in present present in match with the description. in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bbd61945af33e32c336d956e4144196ab9a52b0a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0dd370063d6e35cc96a99cbcb851ad5479ce7622b95e6b2e05477366b0d71fd0
+size 129576
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..bf0287c5c814b49b6aee5442626367b9570b6ce7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a residential area where a house has been severely damaged by a fire. The house is surrounded by yellow police tape, indicating that it is off-limits. The fire has caused significant structural damage to the house, with the roof and upper floors visibly destroyed. The house is situated between two other houses, which appear to be undamaged. A car is parked in front of the damaged house, and a staircase leads up to the front door. The sky is clear and blue, suggesting that the fire occurred during the day. The overall style of the video is realistic and unembellished, capturing the aftermath of a tragic event."
+    ],
+    "video_ids": [
+        "RymC2E47YQE_7_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Damaged house, yellow police tape, two undamaged houses, a car parked in front, and a staircase leading to the front door.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a damaged house with visible fire damage, yellow police tape around the perimeter, two undamaged neighboring houses, a car parked in front, and a staircase leading to the front door. All elements described in the condition are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a scene house with a damage damage, yellow police tape indicating the area, two undamaged houses houses, a car parked in front of and a staircase leading to the front door. The elements in in the condition are present in match with the scene content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c683312949b6f9db0357e758bfe4ce1646cf570b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40b5ea853537133d8ce718f8e46723f2b2172a4df53668a61df2622c08fa4724
+size 171448
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..fb08f61ce8cf6e0860959bf56c882dc0225e6264
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the front view of a blue Mitsubishi car in motion. The car's design is sleek and modern, with a prominent logo on the grille. The car is driving on a road with a clear sky in the background. The license plate reads \"EL 2 75\" and has the text \"Beautiful British Columbia\" on it. The car's headlights are on, illuminating the road ahead. The car's hood is slightly raised, suggesting it's in motion. The overall style of the video is dynamic and captures the car's movement and design in a clear and detailed manner."
+    ],
+    "video_ids": [
+        "0IX5oY25DD8_34_90to216"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue Mitsubishi car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a blue Mitsubishi car, identifiable by the Mitsubishi logo on the front grille and the car's distinct styling. The camera pans across the front of the vehicle, revealing details such as the headlights, license plate, and hood, all consistent with a blue Mitsubishi model. No elements contradict the description of the object.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a blue car car, which by its Mitsubishi logo on the front grille and the overall's overall design. The car angle across the front of the vehicle, highlighting its such as the headlights, grille plate, and overall, all consistent with the Mitsubishi Mitsubishi model. There additional contradict the description of a object.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b5cfe8d2452676466bbe1c3b3c838512f5f2447f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a8e91c06e09f83a35b17f5de8e95ece9aa4130d77dac3e3389f5e7f9d7a2e77a
+size 128790
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..8d06549f0b581a42e740b94832a73b4162db0110
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are standing next to a blue car in a parking lot. The man is wearing a blue shirt and sunglasses, while the woman is wearing a white jacket. They are both pointing at the car, indicating that they are discussing something about it. The parking lot is filled with other cars, and there are trees in the background. The sky is clear and blue, suggesting that it is a sunny day. The overall style of the video is casual and informal, with the focus being on the interaction between the man and the woman."
+    ],
+    "video_ids": [
+        "Mh9v7zludUs_28_36to162"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman, a blue car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman standing next to a blue car, which matches the core description. The presence of a red car in the background does not contradict the description, as the condition only requires the presence of a man, a woman, and a blue car.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man and a woman standing next to a blue car. which matches the ' description of The presence of additional parking car in the background does not contradict the main as as it focus only specifies the presence of a blue, a woman, and a blue car.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bb31d4690ffe4b0a6c3e68958b167ccead6090ec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b3b5a2acf660c02d4594935644a29f3fb2df7f2ca09ef583e50f3e50beb6e480
+size 133202
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..3ce63bf720b928113d9b45b269b3ba169f5e7419
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a man in a suit, who appears to be a news anchor or reporter. He is speaking about an NFL investigation, specifically mentioning Roger Goodell and the Patriots. The background includes a brick wall and a framed jersey, suggesting a sports-related context. The style of the video is typical of a news broadcast, with a focus on the reporter and the topic at hand. The video likely includes additional footage or graphics related to the NFL investigation, but the specific details are not visible in the provided frames."
+    ],
+    "video_ids": [
+        "6aigBJL2ENg_1_0to145"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit (news anchor/reporter)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man dressed in a suit, speaking in what appears to be a news studio setting, consistent with the role of a news anchor or reporter. The background elements, such as the 'GET UP' sign and sports memorabilia, support the context of a sports news program, which aligns with the description of a news anchor/reporter discussing sports-related topics.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man dressed in a suit, which in what appears to be a news studio setting. which with the role of a news anchor or reporter. The attire includes, such as the brickTTY' sign and the-relatedabilia, further the context of a news news broadcast. which iss with the typical of a news anchor orer.\"\n sports-related content.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..758e14a9909b3a81f38ef9b023e83c43576bc5fe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dde8ac11f858260dda2b87772b5c85ae630ab099a5ae6f4d2878970b60a6e65e
+size 149933
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..c1d0f4546a8e4600cb1fd20142cc46e57fbee416
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a dynamic and stylish advertisement for a red BMW sports car. The car is captured in motion on a race track, showcasing its sleek design and powerful performance. The car's vibrant red color stands out against the backdrop of the track, which is lined with trees and a fence. The car's black rims and black roof add to its sporty and aggressive look. The video is shot from a low angle, emphasizing the car's low and wide stance. The car's license plate reads \"K8 087\". The video is likely to be used for promotional purposes, highlighting the car's speed, agility, and luxury."
+    ],
+    "video_ids": [
+        "KhJW8g-BgTo_14_41to181"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red BMW sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red BMW sports car, specifically a BMW M2, driving on a racetrack. The car's color, model, and setting align perfectly with the description of a red BMW sports car, and no elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a red BMW sports car, which a BMW M4, which on a racetrack. The car's design, model, and the align with with the description of a red BMW sports car. fulfilling there conflicting contradict this description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6201913d0f315e830514a8806c2ea2b89efa8d6b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8de8879ec67ae7fc69669308d0c529cf5dd450784592738beb24ef9bc72b5651
+size 217969
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..86e959ce3a6b51f8e3d127eecb45ef825ae8170c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a car's wheel in three different stages of motion. The style of the video is a time-lapse or stop-motion animation, capturing the wheel's movement from a stationary position to a rolling motion. The wheel is silver with a distinctive design, featuring a logo in the center. The car's body is white, and the wheel is positioned on the side of the car. The background is blurred, focusing the viewer's attention on the wheel. The video does not contain any text or additional objects. The overall style of the video is simple and straightforward, focusing on the wheel's motion and design."
+    ],
+    "video_ids": [
+        "ICMSnedTBQc_27_0to147"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A silver car wheel with a central logo, a white car body, and the wheel positioned on the side of the car.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a silver car wheel with a central logo, positioned on the side of a white car body. The camera pans out to reveal the surrounding car body, which is white and matches the description. No conflicting elements are present that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showing showing a silver car wheel with a central logo, a on the side of a white car body. The wheel angle to to reveal the full area body, which is white, matches the description. The additional elements are present in would the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..eefc80897eeca0a6e442f2893d513f786f80a6a5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5c1196ff006ddd25c19e738471e1f776b01254c81b0fdc1921af7cecfca9917
+size 126822
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..6f442b2ce2105430c1fd10b8c51d67588513304f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a news anchor named Keith Cate. The style of the video is a professional news broadcast with a focus on a specific topic. The news anchor is dressed in a suit and tie, and he is standing in front of a news desk with a backdrop that includes a large screen displaying a news headline. The headline reads \"CEMINOLE HEIGHTS MURDERS DEATH PENALTY DECISION\". The video likely includes a discussion or report on a recent decision related to the Ceminoole Heights murders. The overall tone of the video is serious and informative, typical of a news broadcast."
+    ],
+    "video_ids": [
+        "lMYw2ed7A_0_0_0to158"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: News anchor named Keith Cate, news desk, large screen backdrop with a headline.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a news anchor, identified as Keith Cate, sitting at a news desk. Behind him is a large screen displaying a headline about the 'SEMINOLE HEIGHTS MURDERS DEATH PENALTY DECISION'. All core elements of the 'Object(s)' condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a person anchor in presumably as Keith Cate, standing at a news desk. The him, a large screen backdrop a headline, PhD EuropeanEFCAR'S'URDERERS.'BSALTY CASEISION.' The elements elements of the 'Object(s)' condition are present: match described in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1be2f5ec744033b34b4a9d3ef03136ef414affb5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1258-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eeefd54e0e7b6cbebe9940eceb69f98ea4d3eff24d3bc274f963a1eb8594ddcf
+size 118564
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..b8bc0a8764e8e0232b88879a90bdaa9a66723612
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a black Mercedes-Benz car driving on a snow-covered road. The car is sleek and modern, with a shiny black exterior that reflects the surrounding environment. The road is covered in a layer of snow, and the car's tires leave tracks as it moves forward. The car is driving towards the camera, and the angle of the shot suggests that the car is moving at a moderate speed. The background is a mix of trees and houses, indicating that the car is in a suburban or rural area. The overall style of the video is realistic and naturalistic, with a focus on the car and its movement through the snowy landscape."
+    ],
+    "video_ids": [
+        "FlO_6Cy8EFA_6_105to233"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black Mercedes-Benz car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black Mercedes-Benz car parked on a snowy surface, which matches the description. The car's design, emblem, and color are consistent with a Mercedes-Benz model, and there are no elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black Mercedes-Benz car, on a snowy street. which align the description of The car is design, including, and overall are consistent with the Mercedes-Benz.. and the are no additional in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..43a32c166ad4f32652de6150877648b98f37e511
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d15dedfaba246f29e6bb7e0f42b5341265e380c45a01baaf191e2cec064b4c1a
+size 189635
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..ca460a12616d4919b75a9d354037c7a63ce85999
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dramatic scene set in an open field under a cloudy sky. The focal point is a wooden post with a yellow strap around its middle, standing upright on a wooden pallet. Atop the post sits a black bird-like figure, seemingly perched and ready for action. Suddenly, a black ball is launched towards the bird from the left side of the frame. As it approaches, the bird remains stationary, creating a moment of anticipation.  Upon impact, the bird is propelled backward, and a large cloud of white powder or smoke erupts from the base of the post, obscuring the view momentarily. The bird is then seen flying away to the right, leaving behind a trail of debris and smoke. The background features a vast expanse of grassy terrain with some distant hills and a few scattered trees, adding depth to the scene. The overall atmosphere is one of sudden action and explosive energy, captured through a single, dynamic shot."
+    ],
+    "video_ids": [
+        "e903334d439ab4d6d180bfb5ab883c0daf750f870d189d7ce595e39dffd925a0"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A wooden post with a yellow strap, a black bird-like figure, a black ball, and a large cloud of white powder or smoke.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a wooden post with a yellow strap, a black bird-like figure perched on top, a black ball in motion, and a large cloud of white powder or smoke resulting from an explosion or impact. All described elements are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a wooden post with a yellow strap, a black bird-like figure perched on top of and black ball ( the, and a large cloud of white powder or smoke. from the explosion. similar. The these elements are present and match with the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e300a388e127cffcddf659a3ce868d5d53c6ec8b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc49ad6ae94f4ca06dafa710975b4c3c054cbd30fb64975293a333050605a257
+size 189592
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..e2d3cf2ad8916038ac6d9397e03fa5f69f5b9f36
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a red car's engine compartment. The engine is visible with various components such as the air filter, spark plugs, and fuel lines. The car's hood is open, revealing the engine's intricate details. The style of the video is a time-lapse or a series of still images, capturing the different parts of the engine in a sequential manner. The focus is on the engine's components and their arrangement, providing a detailed view of the car's internal workings. The video does not include any text or additional elements, keeping the viewer's attention solely on the engine."
+    ],
+    "video_ids": [
+        "wbTl8RXvONE_155_16to144"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Engine, air filter, spark plugs, fuel lines? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a detailed view of a red car's engine, including components such as the air filter (the silver rectangular cover with circular openings), spark plugs (the cylindrical components on the engine cover), and fuel lines (the yellow hose and other metallic tubes connected to the engine). All specified elements are present and visible, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts an car view of a car car's engine compartment which components such as the air filter,visible black box part), a patterns), spark plugs (the red components with the top block), and fuel lines (the purple and visible purple connected components). to the engine). The these objects are present and identifiable in fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e2e9364afc8c7cebcb8c0862624099e414a285af
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1df481cfed3e8ba0ad45316b483b4fd6f88910d2bc95ca72bfe374fac88827e4
+size 105910
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..0b6749cbd08c6c1bf812a20779ad6294e0c62c64
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a bustling city street filled with various vehicles, including cars and trucks, all moving in the same direction. The street is lined with trees and traffic lights, adding to the urban atmosphere. The vehicles are of different colors and models, suggesting a diverse group of drivers. The street appears to be busy, with traffic moving smoothly under the guidance of the traffic lights. The overall style of the video is realistic, capturing a typical day in a busy city."
+    ],
+    "video_ids": [
+        "cpmRL1cZCHY_29_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Various vehicles including cars and trucks? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple vehicles including cars and trucks moving along a city street. Various models and colors of cars are visible, including sedans, SUVs, and a convertible, which aligns with the 'Various vehicles including cars and trucks' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a vehicles, cars and trucks, through a road street. The types of types of cars are visible, and sedans and SUVs, and a truck truck which aligns with the descriptionVarious vehicles including cars and trucks' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d4da0d11f667a89203b2fb99a7050fd9f6a420fe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d298d12017df3559aa1ab86b972ba3dbfb25f06fdb839d71b3135a65a325f782
+size 242469
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..07af8838c59772782a8f1bfdf4e753444c5ae772
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the growth of a pumpkin plant in a garden. The first frame shows a small pumpkin beginning to form, surrounded by large green leaves. In the second frame, the pumpkin has grown larger and is now nestled among the leaves, which are starting to show signs of decay. By the third frame, the pumpkin has reached its full size and is ready for harvest, while the leaves have turned brown and withered, indicating the end of the growing season. The video is a time-lapse that beautifully illustrates the transformation of a pumpkin plant from seedling to harvest-ready fruit."
+    ],
+    "video_ids": [
+        "8vjtlcMwAGk_4_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Small pumpkin, large green leaves, fully grown pumpkin, decaying leaves.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small, light-colored pumpkin nestled among large green leaves, some of which are decaying or showing signs of wilting and browning. The pumpkin appears to be fully grown, and the surrounding foliage matches the described elements. There are no conflicting elements in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a small pumpkin orange orange pumpkin surrounded among large, leaves, which of which are fullyaying. turning signs of agingting. turningrowning. The sequence appears to be in grown, as the leaves leaves includes the description '. The are no additional elements in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b964742147d0a63432c67e571ab60c95446a4d36
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb615b63170cd02f1fb3578168f5e1d4daf6f0aee7176d8e579b2374945ff988
+size 293307
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..75754493264aeb50469f8005406f16ae8db26252
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a scene set on a wooden desk in what appears to be an indoor environment, likely a home office or study area. The desk is cluttered with various items, including a computer monitor displaying a desktop background featuring a bird, a black LG monitor, a keyboard, a mouse pad with the text \"The Rother,\" a computer mouse, a small bell with a green button, a bottle of lotion, and a pen holder containing pens and pencils.  A pigeon, predominantly white with gray and black markings on its wings and tail, enters the frame from the left side. It walks across the desk, passing by the bell and the computer monitor. The pigeon's movements are deliberate and curious as it explores the desk space. At one point, the pigeon interacts with the bell, pecking at the green button, which suggests a playful or inquisitive behavior. The pigeon continues to move around the desk, occasionally stopping to look around, before eventually exiting the frame towards the right side"
+    ],
+    "video_ids": [
+        "e693cd6fe78ca9d02eaa2d34ecea7cc26184b1da6bca9f8c96a6507f62cfb717"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Computer monitor, black LG monitor, keyboard, mouse pad with 'The Rother' text, computer mouse, small bell with a green button, bottle of lotion, pen holder with pens and pencils, pigeon? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a computer monitor (LG brand), keyboard, mouse pad with 'The Rother' text, computer mouse, small bell with a green button, bottle of lotion, pen holder with pens and pencils, and a pigeon on a wooden desk. All described objects are present and visible in the video without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a computer monitor,part brand), a, mouse pad with textThe Rother' text, computer mouse, small bell with a green button, pen of lotion, pen holder with pens and pencils, and a pigeon. a desk desk. The elements elements are present and accurately in the video.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..33f6937d07d969553f5e7c95d9e422b56056843e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28ecb2918054f3a79d117c372e78ba10fae15892569a35960f4384c374ca00b3
+size 169606
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..86e7187197dc5ad516913dc8cc017ffca7638e97
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a split-screen, showing a young woman on the right and a music streaming interface on the left. The woman is smiling and appears to be enjoying the music. She has long brown hair and is wearing a black tank top. The music streaming interface displays the song \"Roses - Imanbek Remix\" by SAINt JHN, Imanbek. The song is paused at 2:56. The style of the video is casual and personal, capturing a moment of the woman's life while she listens to music."
+    ],
+    "video_ids": [
+        "L_rTh0cbQ5g_15_50to232"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Young woman, music streaming interface? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly features a young woman speaking to the camera, and a music streaming interface is prominently displayed on the left side of the screen, showing the song 'Roses - Imanbek Remix' by SAINT JHN, Imanbek. The core elements described are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a young woman and, the camera, and there music streaming interface is displayed displayed in the screen side of the screen. fulfilling various interface 'Eum' 'mbek'' by 'Z JHN. whichmanbek, The presence elements of in present, do depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7c9b432246463b49586a05ff96fd6b95a6dee800
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7d9299927b8d196015b3090eaf75367d76e7b2d6c62424b92056199180b4fa0e
+size 107449
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..cc55dd423555b47d5790820b6e584b4071d68eac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a character in a fantasy setting, standing in a snowy landscape with mountains in the background. The character is wearing a dark, armored suit with spikes and a helmet, giving off a menacing and powerful aura. The armor is detailed and intricate, suggesting a high level of craftsmanship. The character is standing with their arms outstretched, as if commanding or controlling something. The overall style of the video is realistic with a touch of fantasy, and the lighting is soft, casting a warm glow on the snowy landscape. The character's pose and the surrounding environment suggest a sense of adventure and exploration."
+    ],
+    "video_ids": [
+        "h6p7da3t1do_1_0to170"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A character in a dark, spiked, and detailed armored suit with a helmet.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a character wearing a dark, spiked, and detailed armored suit with a helmet, which matches the description. The armor is intricately designed with sharp, angular spikes and a helmet that covers the head, fitting the specified criteria.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a character wearing a dark, spiked, and detailed armored suit with a helmet. which align the description provided The armor is blackately designed with spikes spikes sp spikes and a dark that covers the entire, fitting the ' '.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0b0faf705fc29efe8612e4c0ae49e4201cb620b3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1259-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60540cc023dad34d6aac1d2adcb58d537b5c1096f4ac004e418278ff70095cf9
+size 207130
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..d0b6cd7db2aa6910752ccac3ff32e072cd2b3cf4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse of a vintage car's interior, showcasing the steering wheel and dashboard. The steering wheel is white and has a classic design, with a silver center. The dashboard is green and features a radio and gauges. The car's interior is brown, with a leather seat visible in the foreground. The car is parked outside, with trees and a road visible through the windows. The time-lapse effect gives a sense of motion and change, as the car's interior is shown in different lighting conditions throughout the day. The overall style of the video is nostalgic and evocative of a bygone era of automotive design."
+    ],
+    "video_ids": [
+        "BDT39Gglirc_13_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, dashboard, leather seat? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a steering wheel, dashboard, and leather seat, which are the core objects mentioned in the condition. The interior of the vehicle is visible, and these elements are prominently featured throughout the frames. The video does not contradict the description, and additional elements (like the door panel or window view) do not conflict with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a steering wheel, dashboard, and leather seat, which are the main elements mentioned in the description. The presence of the car is depicted, and the elements are prominently displayed, the frames. The presence does not include the description by and the elements suchlike the road and and the)) do not conflict with the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..35e0bef21a922bf3b883b83f1747d2c9a42966c0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c86e8e0149cba924fa2c74b4dbc11b493bfd9ce00d86c496e962c6c57b41a666
+size 129309
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..6ee5174d489c131587bce02b712d416dba879817
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a bald man is seen sitting at a table, engaged in a conversation with another person. He is dressed in a blue jacket and a blue shirt, and his hands are clasped together on the table. The table is adorned with a vase of flowers, adding a touch of elegance to the setting. The man's expression suggests that he is listening intently to the other person, indicating a serious or important discussion. The overall style of the video is realistic, capturing a moment of interaction between two individuals in a formal setting."
+    ],
+    "video_ids": [
+        "X3Tb6IcYSIM_258_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bald man, another person, a table, a vase of flowers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a bald man seated at a table, engaged in conversation with another person (partially visible). A vase with flowers is also visible on the table in the background, fulfilling all the specified object conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a bald man in at a table, engaged in a with another person.partially visible). There vase of flowers is also present on the table, the background. fulfilling all the conditions conditions conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..19d312acaec69c42e7e8a72585422209e3c5ad0f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7580224039bf19fe9c316c97e1cc7617cc8972e4c77eb972af676fd39ec382a3
+size 114883
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..e8cfdf25b15559e27e9cd8a32ccbbe2ed2e52438
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a yellow sports car driving on a snowy road. The car is sleek and modern, with a shiny finish that reflects the surrounding environment. The road is covered in a layer of snow, and the car's tires leave tracks as it moves. The car is the main focus of the video, and it is captured from a side angle, allowing viewers to see its design and features. The background is a barren landscape, with rocky hills and sparse vegetation. The sky is overcast, and the overall atmosphere is cold and desolate. The car's bright color stands out against the muted tones of the landscape, creating a striking contrast. The video has a cinematic quality, with a focus on the car and the road, and a sense of motion and adventure."
+    ],
+    "video_ids": [
+        "qyxIvTzSqJA_1_0to103"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A yellow sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a yellow sports car, specifically a Jaguar F-Type, which matches the description. The car is the central focus of the video, and no elements contradict the core description of a yellow sports car.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a yellow sports car, which a convertible X-Type, which matches the description of The car is the central focus of the video, and its other contradict the core description of the yellow sports car.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2ad76b5cacc9b6235cdf0d60a4f0a4d1fa062d5d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b9865bf9c574b9fce8fb7bab236018aaa10c14a0f849a4a5a627dce61c71ae30
+size 269284
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..6a36f17d83f7699b0a89cda1d78d88677ada147f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a man in a suit being interviewed by a group of reporters. The man is wearing glasses and appears to be speaking into a microphone. The reporters are holding up their microphones towards the man, indicating that they are recording his statements. The setting appears to be outdoors, as suggested by the natural light and the presence of a fence in the background. The style of the video is a standard news report, with the focus on the man being interviewed and the reporters capturing his words. The video likely contains audio of the man speaking and possibly the questions posed by the reporters."
+    ],
+    "video_ids": [
+        "OQCSXdVri3U_5_0to143"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit wearing glasses, a group of reporters holding microphones.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit wearing glasses, surrounded by reporters holding microphones. The scene matches the described 'Object(s)' condition, with no conflicting elements. Additional elements such as smartphones and other reporters are present but do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man in a suit wearing glasses, which by multiple holding microphones. This core is the description elementsObject(s)' condition without with the additional elements present The elements like as the and the reporters' present but do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..09dc4fe5ddb532b7073be74cca54be8f6ef840f6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:51d9ad23e82febab3c878b107647903f91e84e505ae1b16c122fd6aa23ffabed
+size 148713
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..c3a91910590ad86812c44021f57fd924745cfe89
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse aerial shot of a mountainous landscape with a large body of water. The first frame shows a clear blue sky with a few clouds, and the landscape is lush and green with trees and vegetation. The second frame shows a boat moving across the water, leaving a wake behind it. The third frame shows the landscape with the boat still moving, but the sky has become overcast with clouds. The style of the video is a realistic aerial shot, capturing the beauty of nature and the movement of the boat across the water."
+    ],
+    "video_ids": [
+        "_DWeSm9YtpU_90_316to446"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Boat, trees, vegetation, mountains, water, sky, clouds? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a boat moving across the water, surrounded by trees and vegetation on numerous islands and hills, with mountains in the background and a sky with clouds visible above. All specified elements are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a boat, across a water, surrounded by trees and vegetation on the hills. hills. with mountains in the background. a sky with clouds above in the The the elements are present and contribute depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..38ddce524d855661826d64606e17c784fb66e105
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0e2a56eec0c41d087dc207193838077c7afdd4b373149a1eab227204a82c03f
+size 155330
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..c41213f0899b6cfa9b5a206dbd87fcf9fa7273c7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman in a white hat and black tank top is seen walking a brown horse down a gravel path. She is followed by two other women, one of whom is holding a child's hand. The path is lined with tall grass and trees, creating a serene and peaceful atmosphere. The women are dressed casually, and the horse is calm and well-behaved. The overall style of the video is natural and organic, capturing a simple yet beautiful moment in time."
+    ],
+    "video_ids": [
+        "EC8Rv1MsKeM_52_0to147"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a white hat and black tank top walking a brown horse, followed by two other women with one holding a child's hand.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a white hat and black tank top walking a brown horse, and behind her are two other women, one of whom is holding a child's hand. This matches the core description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a woman in a white hat and a tank top riding a brown horse. which behind her, two other women, one of whom is holding a child's hand. The matches the description description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1058141f073848c18fd7d645ce421c5c92859527
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c2d14fa0451bcb3ca72624405f47b7b9e6278263d52ad064c5b40a721df765dc
+size 174462
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..c750b849bcf2bad3c8302dddd040a7c68325537b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a silver car parked on a street. The car is facing the camera, and its headlights are on. The car is parked next to a grassy area with trees. The sky is blue and clear. The car is the main focus of the video, and it is stationary. The background features a grassy area with trees, and the sky is clear. The car is parked on the street, and there are no other vehicles or people visible in the video. The car is the only object in motion in the video. The car is parked on the street, and there are no other vehicles or people visible in the video. The car is the only object in motion in the video. The car is parked on the street, and there are no other vehicles or people visible in the video. The car is the only object in motion in the video. The car is parked on the street, and there are no other vehicles or people visible in the video. The car is the only object in motion in the video. The car is parked on the street, and there are no other vehicles or people visible in the video. The car is the only object in motion in the video. The car is parked on the street, and there are no other vehicles or people visible in the video. The car is the only object in motion in the video. The car is parked on the street, and there are no other vehicles or people visible in the video. The car is the"
+    ],
+    "video_ids": [
+        "LlnKnJr1nJM_41_0to117"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Silver car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a silver car, with clear shots of its front bumper, headlights, and wheel. The car's color and model are consistent with the description, and no elements contradict the presence of a silver car.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a silver car, which its visibility of its front and, grille, and grille, The car is color is design are consistent with the description of and there other contradict the ' of a silver car.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2d94eb50b6622d602e660e4309a8b6ce8aca437f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:334727a901378d84f2e0ef978f744fc9e5a237a9b61a9bfb29f6192d6d87f905
+size 219851
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..167177d347a0eb14b59af987245de7947bc91cbb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a vibrant and colorful display of fresh fruits and vegetables. The scene is set on a table, where a variety of produce is neatly arranged. The fruits and vegetables are in focus, with the camera capturing their textures and colors in detail. The fruits include oranges, bananas, and apples, while the vegetables include broccoli, tomatoes, and peppers. The arrangement of the produce suggests a healthy and balanced diet. The video is likely to be used for promoting healthy eating habits or showcasing the variety of produce available at a market or grocery store. The style of the video is realistic and naturalistic, with no artificial elements or filters applied. The colors are bright and vivid, reflecting the freshness of the produce. The lighting is natural, with no artificial lighting used. The overall impression is one of abundance and variety, inviting viewers to appreciate the beauty and diversity of nature's bounty."
+    ],
+    "video_ids": [
+        "RwGLQ-aM8o4_48_0to134"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Oranges, bananas, apples, broccoli, tomatoes, peppers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows oranges, tomatoes, and peppers as requested. Bananas are visible in the background, and broccoli is partially visible on the left side, wrapped in plastic. Although apples are not clearly visible, the presence of other fruits and vegetables does not contradict the description, as the core requested items are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a, bananas, and peppers, described. Whileanas and also in the background, and there is not seen in the right side. which in a. The there are not clearly visible, the presence of other fruits like vegetables suggests not contradict the core. as the core objects objects ( present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..09bb029a30d945a1557eea6755b371f39a5af2bd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-126-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ce86e8bfc6e7a628375c18f65cc5713a3369001d127966145500d14046e656c
+size 130025
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..c71210fd2c097c1bec10cdc9a9e45674105c19a2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a pair of silver and black headphones with a black cable, a pink iPhone, and a black earbud with a gold tip. The headphones are placed on a wooden table, and the iPhone is lying next to them. The earbud is connected to the iPhone, suggesting that someone is listening to music or a call. The style of the video is simple and straightforward, focusing on the objects without any additional context or action. The colors are muted, with the silver and black of the headphones contrasting against the pink of the iPhone and the black of the earbud. The wooden table provides a natural and warm background for the objects. The video does not contain any text or additional elements, keeping the focus solely on the headphones, iPhone, and earbud."
+    ],
+    "video_ids": [
+        "noE8RvEw1nk_13_31to173"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: silver and black headphones, black cable, pink iPhone, black earbud with a gold tip? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows silver and black headphones with a black cable, a pink iPhone, and a black earbud with a gold tip. All described objects are clearly visible and match the description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a and black headphones, a black cable, a pink iPhone, and a black earbud with a gold tip. These the elements are present visible and match the description provided any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..424b4be0331459a2e5e8e61130294821a6781d40
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c17ac3121da704000b319c94b8d02896b03beec575c8d58772bd3960ee9c9d6
+size 52693
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..31ec4d29fcf0ad2542007d587b1439cefa3a514e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated sequence featuring three anthropomorphic cat characters in a room with a wooden floor and a pink wall. The first cat, a light brown tabby, walks with a confident stride, its tail held high. The second cat, a white cat with a black nose, follows closely behind, its eyes focused on the tabby. The third cat, a pink cat with a white muzzle, walks in the opposite direction, its eyes closed as if it is lost in thought. The room is filled with various objects, including a yellow shelf filled with colorful items and a pink window with a white curtain. The overall style of the video is whimsical and playful, with a focus on the expressive faces and movements of the cat characters."
+    ],
+    "video_ids": [
+        "z_OlG23s-JU_3_60to192"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ['Three anthropomorphic cat characters', 'Yellow shelf', 'Pink window', 'White curtain']? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows three anthropomorphic cat characters (brown, white, and pink), a yellow shelf in the background, a pink window with visible frames, and a pink curtain hanging beside the window. All specified objects are present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts three anthropomorphic cat characters,one, white, and pink), a yellow shelf, the background, and pink window, a green, and a white wall. on the window. The these elements are present and match the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..630c1ac93402c28ec95093a402e1865db5e85629
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c48b53b66f73190c13613f024719a6f715807bf1be63b853f0690b6778db3316
+size 178664
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..064c3636b3ab840a7bc3141739b1181a2cd4ae39
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene at a marina. The main focus is a large white yacht, which is docked and appears to be the center of attention. The yacht is surrounded by several smaller boats, creating a sense of depth and scale. The water is calm, reflecting the light and adding to the tranquil atmosphere. In the background, there are buildings, suggesting that the marina is located near a city or town. The overall style of the video is aerial, providing a bird's eye view of the marina and its surroundings. This perspective allows for a comprehensive view of the scene, highlighting the yacht and the smaller boats, as well as the buildings in the distance. The video is likely to be used for promotional purposes, showcasing the marina and its facilities to potential visitors or customers."
+    ],
+    "video_ids": [
+        "EvBm1h3_tF0_0_0to139"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large white yacht, several smaller boats, buildings in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a large white yacht as the central focus, surrounded by several smaller boats including sailboats and inflatable dinghies. In the background, there are visible buildings along the shoreline, consistent with the described elements. The scene is well-composed to match the specified objects without contradicting them.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a large white yacht at the central object, surrounded by several smaller boats. sailboats and motor rahies. In the background, there are buildings buildings, the shoreline, which with the description elements. The scene is set-composed, fulfill the given conditions.\"\n anying the.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..79fd7b88e1f3ad17c2de499aa995ee31b9a6c3a2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c070ffac960dcca70c5b0752c7f7c9a2de961a7c1452ea3110cc53201f03424b
+size 130288
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..7738f7df313933c734435a7d95c4ea02393a39f8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a vibrant orange BMW car parked in a parking lot. The car is positioned at an angle, showcasing its sleek design and shiny exterior. The car's black rims contrast with the orange body, adding to its sporty appearance. The parking lot is mostly empty, with only a few other cars visible in the background. The sky above is cloudy, suggesting an overcast day. The overall style of the video is dynamic, with the car being the main focus, and the surrounding environment providing context. The video captures the essence of the car's design and color, making it a visually appealing piece."
+    ],
+    "video_ids": [
+        "QAVwcK-OkcQ_20_19to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A vibrant orange BMW car parked at an angle, showcasing its sleek design and shiny exterior. The car has black rims.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by showing a vibrant orange BMW car parked at an angle, with its sleek design and shiny exterior clearly visible. The car also has black rims, matching the description. Additional elements in the background (like parking lot markings and distant structures) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition by depicting a vibrant orange BMW car parked at an angle. which a sleek design and shiny exterior prominently visible. The car's has black rims, which the description. The elements like the video,like the lines lines and a objects) do not conflict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..67fadd4fc7fe059f55293203df74cff82e52aef1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:caf88c5e162135bc377ffcebbf877ead3591ff4f81aac3a6e2b7139913abe90f
+size 360507
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..0422408f671b63a6aacc9b727e2afd92cbcf686a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic scene at a skate park. A red BMX bike is the main focus, parked on a ramp. The bike is equipped with a black seat and handlebars, and it's positioned in the center of the frame. The ramp, made of concrete, is located in the background, adding depth to the scene. The skate park itself is filled with various ramps and obstacles, suggesting a space designed for action and stunts. The overall style of the video is energetic and vibrant, capturing the essence of the skate park and the thrill of BMX biking."
+    ],
+    "video_ids": [
+        "t12GSql7DWM_18_0to152"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red BMX bike? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red BMX bike as the central object, shown in various shots including a full view and close-ups of its wheel. The bike's red frame and BMX characteristics are clearly visible, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a red BMX bike, the main object, which in a angles from close close view and a-ups. the details and The bike is design color and blackX design are clearly visible, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6172bc5d42b56b102114ecb7356fceccf234a78c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:43133c678f5004467fb2c85d9c79cd99eade88360388f2370cd683b1f2a6f6bb
+size 161175
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..b9a01c07e83bac4b6fd86c9951f1df0faed62cd1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a short, artistic piece featuring a man holding a painting of a face. The man is dressed in a plaid shirt and is seated in a red chair. The painting is a striking portrait of a man with a large nose and a small mouth, rendered in bold, contrasting colors. The man holding the painting appears to be in deep thought, perhaps contemplating the meaning of the artwork. The overall style of the video is minimalist and modern, with a focus on the interplay between the man, the painting, and the red chair. The use of color and light creates a dramatic effect, highlighting the emotional intensity of the scene."
+    ],
+    "video_ids": [
+        "CGLd-f6ir3s_13_0to140"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a painting of a face, a red chair? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man (partially visible on the left and right sides), a painting of a face (central focus, cubist style), and a red chair (visible in the background). These elements are consistently present throughout the video, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man holdingtheially visible) the left side right sides of a painting of a face (center focus), heldist style), and a red chair (the in the background). The elements match consistent present throughout the frames, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a583b2ecf7e3683870c1c5e2c5cc9876c6b10124
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a51a3e0955b465585ee3f3ab6bc522e74665079c4a504b6066ea8651f4a2e925
+size 116418
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..0724fffe04323466e85e02e4eb45a7ce70536123
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a gray polo shirt and glasses, sitting in a room with a bookshelf in the background. He is gesturing with his hands, pointing at something off-camera. The room has a casual and lived-in feel, with various objects scattered around, including books and a sports ball. The man appears to be engaged in a conversation or presentation, as he is actively gesturing and speaking. The overall style of the video is informal and personal, suggesting a relaxed and comfortable setting."
+    ],
+    "video_ids": [
+        "D3QznppVuGU_2_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a gray polo shirt and glasses? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a gray polo shirt and glasses, which matches the specified description. The background elements, such as shelves and posters, do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a man wearing a gray polo shirt and glasses. sitting matches the description '. The man,, such as the with a, do not contradict the description description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fadf73048bd8672b40f5be645bd6d7cb14da7951
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:12bc11f2d870f40dc3a554aa9729017eebb3a3c012b0942932e5fefbdf030f61
+size 242082
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..cb9a7ae52eca5362ba0393c0dc0bf91306e836a8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a black car, focusing on the driver's side. The car is well-maintained and clean, with a sleek design. The driver's seat is black leather, and the steering wheel is also black. The dashboard is equipped with a touch screen display, and there are various buttons and controls on the steering wheel and the center console. The car appears to be parked in a garage or a driveway, as there is a green hedge visible through the window. The video is a straightforward, unembellished depiction of the car's interior, with no additional action or movement."
+    ],
+    "video_ids": [
+        "KO60sbcy9mg_8_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Driver's seat (black leather), steering wheel (black), touch screen display on the dashboard, buttons and controls on the steering wheel and center console? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a black leather driver's seat, a black steering wheel with visible buttons and controls, and a touch screen display on the dashboard. The center console also features various buttons and controls. All described elements are present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a car leather driver's seat, a black steering wheel, a controls and controls, and a touch screen display on the dashboard. The interior console also appears buttons buttons and controls, The these elements are present and match depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..625315f6efaa828c79f57138ac877ef86018047c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1260-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e6779cace1494cee13ff58564b78e8841bd694d2c2547ed45d6d124ce982ebd9
+size 160508
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..07a7eab19210e71dbad9b3180383bd88cc10ff07
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a blue sports car in a studio setting. The car is sleek and modern, with a shiny exterior that reflects the studio lights. The car is positioned on a white platform, which contrasts with the car's blue color. The car's design features include a large grille, sharp headlights, and black rims. The car is stationary, and there are no other objects in the frame. The style of the video is a professional car advertisement, with a focus on the car's design and features. The lighting and composition are designed to highlight the car's sleek lines and modern design."
+    ],
+    "video_ids": [
+        "MqAHGcuXvFM_6_0to134"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a blue sports car, which is the central focus. The car's sleek design, vibrant blue color, and sporty aesthetics align perfectly with the description of a blue sports car. There are no conflicting elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a blue sports car, which matches the main object. The car's design design, vibrant blue color, and sporty features align with with the description of a sports sports car. There are no additional elements in would the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c9438d0f9d57373275e5a85b6ca7882c27c7dd74
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:215a8298b986d52287bf19eab851d124b0bd60dbf2476a4d3a42d826686b9110
+size 87428
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..d29b7602597104a69463382c728af7d0b0d1dd79
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen preparing a drink in a large silver bowl. The person is wearing a black vest and white shirt, and is holding a green leaf in their hand. They are carefully placing the leaf into the bowl, which contains a clear liquid. The bowl is placed on a wooden table, and there are other items on the table, including a bottle and a teapot. The scene suggests a calm and relaxed atmosphere, with the person taking their time to prepare the drink. The focus of the video is on the person and their actions, with the background being relatively simple and uncluttered. The style of the video is straightforward and documentary-like, with no additional effects or music."
+    ],
+    "video_ids": [
+        "bm8NMU7usks_13_82to206"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a large silver bowl, a black vest, a white shirt, a green leaf, a bottle, a teapot? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person wearing a black vest and white shirt, interacting with a large silver bowl. The person is handling a green leaf and appears to be adding it to the bowl. A bottle and a teapot are also visible on the table. All specified objects are present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person wearing a black vest and a shirt, interacting with a large silver bowl. The person is seen a green leaf and a to be placing it to the bowl, There bottle is a teapot are also visible in the table, The the objects are present and the the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3e26f78c484a6bfaaad6dacfcbb9a7affa50ac5b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06f732dfd9daf6ca8b2ecdf7570be7a65b7fcebcfa108e3c0a325ac63b559cc7
+size 142274
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..fbb34a4b10d85cc5eda5b59aea2617d72fb64784
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man giving a speech or presentation. He is wearing a white shirt and a red tie, and he has a microphone attached to his ear. The man is standing in front of a black background with white text. The text is not clearly legible in the image. The man appears to be speaking with confidence and authority. The style of the video is straightforward and professional, focusing on the speaker and his message."
+    ],
+    "video_ids": [
+        "0Ep32C-3sys_2_0to135"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a white shirt and red tie with a microphone attached to his ear.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a white shirt and red tie, with a microphone visibly attached to his ear. The core description is accurately represented, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a man wearing a white shirt and a tie, holding a microphone in attached to his ear. The man elements is accurately represented in and there additional elements are present.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..89c893026f9c8423fedab947d1d5d95577f376ea
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e9b5dafac83a461975829b2f0e70fa576f898facad0b51e0ac40b790a07b4f5e
+size 125827
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..2e3f9064c3082a196bb106b682ab45aea899b716
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a delightful process of cookie decoration. It begins with a tray of plain, round cookies. In the second frame, the cookies are transformed into adorable owl shapes. The transformation is completed in the third frame, where the cookies are adorned with chocolate chips for eyes and a small piece of candy for a beak, giving them a charming, whimsical appearance. The video captures the joy of baking and the creativity involved in cookie decoration."
+    ],
+    "video_ids": [
+        "JQba-EX8mmQ_26_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Cookies, chocolate chips, candy pieces? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows owl-shaped cookies decorated with candy eyes and a small yellow candy beak, which matches the description of cookies and candy pieces. Chocolate chips are not visible, but the description does not require them to be present for the condition to be fulfilled.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows cookies-shaped cookies, with chocolate pieces and a candy candy nose beak. which align the ' of cookies. candy pieces. The is are not present in but the cookies does not require their to be present. the ' to be fulfilled.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9d4a711d9960a47fad17e1e0d4b78f3a9fb8405b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e2264987aee74a7be11af0e56e782d758cffe824c96655bafafbbe67488c9f7c
+size 105976
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..780855974bbdaf7ca0e072ee3d72e09c524de4eb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a blue shirt standing on a wooden deck with a lake and trees in the background. He is gesturing with his hands and appears to be speaking. The man is the main subject of the video, and his actions suggest that he is engaged in a conversation or explaining something. The setting is serene and natural, with the lake and trees providing a peaceful backdrop. The wooden deck suggests a residential or recreational area. The man's attire is casual, and his expression is serious, indicating that the content of the video may be of a serious or informative nature."
+    ],
+    "video_ids": [
+        "9Scj14AK4VE_32_39to161"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue shirt standing on a wooden deck, gesturing and speaking.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a blue shirt sitting on a wooden deck, gesturing with his hands and speaking. Although he is seated rather than standing, the core description is largely fulfilled as he is actively gesturing and speaking while on the wooden deck. The presence of a lake and surrounding greenery does not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue shirt standing on a wooden deck, gesturing with his hands, speaking. The the is not rather than standing, the description elements of largely fulfilled as the is in gesturing and speaking in on a deck deck. The presence of the lake and trees greenery in not contradict the description.\"\n}<|im_end|>\nGuidId",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9fee995b3cd48df5e428bd3c4af29efea35a0db1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:527e9654d44ae6fdf26131b62daef161cd301766e318273f13b9b0eb59bc42fd
+size 217477
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..7853146f87356d844cd568cfcdc09848b29784ef
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman standing in a wine store, surrounded by numerous bottles of wine. She is wearing a beige sweater and a pearl necklace. The woman is identified as Lettie Teague, a Wall Street Journal wine columnist. The store has a warm and inviting atmosphere, with wooden shelves and a variety of wine bottles on display. The woman appears to be speaking or presenting, possibly discussing wine recommendations or sharing her expertise on the subject. The overall style of the video is informative and engaging, with a focus on the woman's expertise and the inviting ambiance of the wine store."
+    ],
+    "video_ids": [
+        "Dsahb-HxpvM_0_0to190"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman (Lettie Teague), beige sweater, pearl necklace, wine bottles? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman identified as Lettie Teague, wearing a beige sweater and a pearl necklace, standing in a wine store surrounded by wine bottles. The text overlay confirms her identity and role, and the setting matches the description. There are no significant contradictions with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing as Lettie Teague, wearing a beige sweater and a pearl necklace. standing in front setting shop surrounded by wine bottles. The description and and the name and the, and the setting matches the description of The are no contradictions contradictions or the given description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..64358c07338f17789ef6f5273389cd0302f093bd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d171a46460a43c30a32f65cbfa129ac80f7e1e718757f5816003dd4a281d5cfa
+size 108770
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..962cd2e9095b6136f9f037b70dd1cd69ccaba370
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a comparison of two cars, the Krimpstang and the Mystic, from the year 1996. The Krimpstang is a red car with a black stripe and a white logo on the side. The Mystic is a green car with a black stripe and a white logo on the side. Both cars are parked on a street with a blurred background. The video is a side-by-side comparison of the two cars, highlighting their similarities and differences. The style of the video is informative, providing viewers with a visual comparison of the two cars."
+    ],
+    "video_ids": [
+        "WhpTaZpkwW4_9_32to232"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Krimpstang (red car with a black stripe and a white logo), Mystic (green car with a black stripe and a white logo)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video correctly identifies and displays the two cars as 'Krimpstang' (red with a black stripe and white logo) and 'Mystic' (dark green with a black stripe and white logo), matching the description provided. The visual elements of each car align with the specified attributes, and no conflicting details are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two displays two two cars as describedKrimpstang' (red car a black stripe and a logo) and 'Mystic' (green green with a black stripe and white logo). which the description provided. The cars elements of the car, with the given characteristics, including there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a75776aec05916ea27d201358d5e77af5b0dba03
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0253b8db5b5ebb9e8addf7e627a8ceb85c967a984093c803833b765639a7d743
+size 77921
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..5a0c4fe913465783baddc2bd98907c51f03bfd8f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a dish of food, likely a dessert, placed on a small, round, blue plate with a floral pattern. The dish is covered with a clear plastic lid, and a spoon is placed next to it. The food appears to be a baked item, possibly a custard or a flan, with a golden-brown crust on top. The plate is placed on a table with a colorful tablecloth, featuring a pattern of squares in various colors. The style of the video is simple and straightforward, focusing on the dish and its immediate surroundings without any additional context or action. The lighting is bright and even, highlighting the textures and colors of the food and the plate."
+    ],
+    "video_ids": [
+        "5UKKC2ooey4_66_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A dish of food (likely a baked item like custard or flan) covered with a clear plastic lid and a spoon next to it.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dish of food, which appears to be a baked custard or flan, covered with a clear plastic lid. A spoon is positioned next to the dish on a decorative plate, matching the description. The visual elements are consistent with the specified object(s) without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dish of food that which appears to be a baked itemard or flan, covered with a clear plastic lid. There spoon is placed next to the dish on the plate plate. matching the description provided The setting elements in consistent with the given condition condition). condition any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8295fc9665c0da854349815247dfa31edb8ffae5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1261-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2d45c2cf4881737ca14a1b3cc82000a8b7efade52dc5022196bc5f00a627949a
+size 251810
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..a64fa4d1ddc635dc746eaf982112936aa07dbf05
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a tranquil forest trail, with a wooden fence on the left side. The trail is surrounded by lush greenery and towering trees, creating a serene and peaceful atmosphere. The sun casts a warm glow on the path, highlighting the texture of the wooden fence and the softness of the grass. The trail leads deeper into the forest, inviting viewers to follow it and explore the natural beauty of the woods. The overall style of the video is calm and serene, capturing the essence of a peaceful forest walk."
+    ],
+    "video_ids": [
+        "qX4sHLNy7Dw_13_0to101"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Wooden fence, lush greenery, towering trees, grass? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a wooden fence running along a dirt path, surrounded by lush greenery and towering trees. Grass is visible on the slopes and ground, matching the described elements. No conflicting elements are present that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a wooden fence on along the path path, with by lush greenery and towering trees. The is also on the ground and the, and the description elements. The additional elements are present, would the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8c9f9557fc365b4f9826c3c09aad152e7abd096b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2795b1fc3bb2bed4587b3bfe89510dfe5f78a63bae7f6251f27768f453095ce
+size 153622
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..536a52992d041a3e2b072cf5663a037172b936e6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a bald man in a colorful patterned shirt, gesturing with his hands as he speaks. He appears to be in a studio setting with a dark background and a logo in the corner. The man's expression is serious, and he seems to be making a point or explaining something. The style of the video is a standard interview or news segment, with a focus on the man's speech and body language."
+    ],
+    "video_ids": [
+        "IKhFdaiJ3mE_5_0to184"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bald man in a colorful patterned shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man wearing a colorful patterned shirt, which matches the description. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a bald man wearing a colorful patterned shirt, which matches the description provided The man is the elements in not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..eb22557f6eec837a95c23810ea0b1b9743367493
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb39194c22e01c690146ada05104a45f723c234e1f47ff634b08a4b00bb636c7
+size 268527
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..fbc4682a00fad70d555284071d7306b2a87581a9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a doctor using a tablet computer. The doctor is wearing a white coat and stethoscope, indicating their profession. The tablet is held in the doctor's hand, which is visible in the foreground. In the background, there is a blue icon of a network or chain of blocks, suggesting a theme of connectivity or data sharing. The style of the video is professional and informative, likely intended to demonstrate the use of technology in healthcare settings. The focus is on the doctor's interaction with the tablet, with the icon serving as a visual metaphor for the exchange of information or data."
+    ],
+    "video_ids": [
+        "gtBPELQxRF0_9_0to112"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A doctor, a tablet computer, a blue icon of a network or chain of blocks? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person in a white coat with a stethoscope, consistent with a doctor. The person is holding and interacting with a tablet computer. A blue icon representing a network or chain of blocks (five connected cubes) appears on the screen, fulfilling the described elements. No conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person wearing a white lab, a stethoscope, which with a doctor. The person is holding and interacting with a tablet computer. Additionally blue icon resembling a network or chain of blocks ispossibly connected circles) appears in the screen of which the ' condition. The additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..88e2070dab071eeb09a92f3a538beafae987967d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:afb6f88c2eba87a7b8891ae7bf1fd83bfcad276fafd3782e1a3ae2295f47bf40
+size 146075
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..a55ab23e318cb5eb25b9db3e92adf96806049050
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a delightful dessert scene. A white plate holds a stack of three golden-brown crepes, each dusted with powdered sugar. The topmost crepe is generously topped with a scoop of vanilla ice cream, garnished with a sprig of mint. The plate is placed on a white tablecloth, and in the background, there's a glass of water and a teapot, suggesting a cozy and relaxed setting. The video is a simple yet elegant depiction of a sweet treat, likely taken in a restaurant or caf\u00e9."
+    ],
+    "video_ids": [
+        "VwIJIln__Zw_17_143to282"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white plate, three crepes, vanilla ice cream, a mint sprig, a glass of water, a teapot? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white plate with three fluffy, golden-yellow crepes dusted with powdered sugar, topped with a scoop of vanilla ice cream and a mint sprig. In the background, a glass of water and a teapot are visible, matching the description. All elements are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white plate with three cre cre golden-b crepes stackeded with powdered sugar. topped with a scoop of vanilla ice cream and a spr sprig. In the background, there glass of water and a teapot are visible, which the description provided The elements are present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e5d167130a2e7ff740ab5dfaeb7fa4264bb7ca4e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2354c268dd6ee5f1d0e1b5c0c27392c08a9b99eea0a2cc612b6e934863900b6
+size 173456
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..fd94200accb4f565bf84a26ea5787c41c21767fa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a split-screen comparison of two men, one shirtless and one wearing a hoodie, set against contrasting backgrounds. The shirtless man is standing in front of a palm tree, showcasing his muscular physique. The man in the hoodie is in a room with shelves filled with various items, including a backpack and a TV. The style of the video is a side-by-side comparison, possibly highlighting the difference in lifestyle or fitness levels between the two individuals. The video could be used for motivational purposes or to compare different lifestyles."
+    ],
+    "video_ids": [
+        "TVR-7eh-GqU_41_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, a palm tree, shelves, backpack, TV? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by showing two men: one shirtless with a muscular build in front of a palm tree, and another man in a hoodie in an indoor setting. The indoor setting includes shelves with various items, a backpack hanging on the wall, and a TV screen (or monitor) in the background. All specified objects are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showing two men, one shirtless and a palm build and the of a palm tree, and the wearing wearing a blue with a indoor setting with The palm setting includes shelves and various items and a backpack, on the back, and a TV..part a) visible the background. The these elements are present, contradiction.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..63418934a6afaad61f7a2261cd75628be3e43d15
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b70c60ff5154f7affa93896b1da34d15b395eedea580a9d6ef4a49a4b7f117ca
+size 270242
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..f9e1d18c1e4c1b62b62b5c5e9f6cfba1a562ce44
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man in a white lab coat stands in a medical office, engaging in a conversation with another man who is standing with his arms raised in the air. The office is equipped with a sink, a chair, and a window. The man in the white lab coat appears to be a doctor, and the other man seems to be a patient. The scene captures a moment of interaction between the two individuals, possibly discussing the patient's health or treatment plan. The office setting suggests a professional and clinical environment."
+    ],
+    "video_ids": [
+        "criOk-C3z-E_37_27to162"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a white lab coat (doctor) and another man with arms raised (patient).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a white lab coat, consistent with a doctor, and another man who raises his arms, matching the description of a patient. The scene is set in what appears to be a medical office, and the actions align with the requested condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man in a white lab coat, which with the doctor, and another man with appears his arms, consistent the description of a patient. The setting is set in a appears to be a medical office, which the actions of with a typical scenario.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1c9bc8a66845b52fcbd70e5b053e1fa1b6f53320
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f6c6f239432c31faf2bc63985f6fb69fba2177caff4545068c3faa99a6a9a0b
+size 172104
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..94a9a7d60adcb42528814bf87504b033bdae4933
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a red baseball jersey sitting at a table filled with various bowls of food. The man is wearing a baseball cap and appears to be in a casual dining setting. The table is covered with a variety of dishes, including bowls of soup, salads, and other food items. The man seems to be enjoying his meal and is pointing at something on the table. The setting suggests a relaxed and comfortable atmosphere, with the man enjoying his meal in a leisurely manner. The video captures the essence of a casual dining experience, with the man enjoying his food and the ambiance of the restaurant."
+    ],
+    "video_ids": [
+        "QdR9sMqnvLU_5_0to145"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a red baseball jersey, wearing a baseball cap, and several bowls of food including soup and salads.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a red baseball jersey with 'Busan' on it and a baseball cap, seated at a table with multiple bowls of food, including what appears to be soup and various side dishes (salads or kimchi). The core elements described are clearly present, and no significant contradictions are observed.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a red baseball jersey and aVch' written the, a baseball cap, which at a table with several bowls of food, including what appears to be soup and salads salads dishes.salads). otherchi). The man elements of in present present, and the additional contradictions are noted.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2f77e1d7769ff772fb74967f812828f57d77d017
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e39c5f3eef8ed467e0c8ee1b0ae3ee8665e6a1a49286a1e2a174e72d45ace67
+size 141039
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..bb4f6a8930ee3671def8345f1a65cd4b6694fa95
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a man in a suit and tie, who appears to be a news anchor or reporter. The man is seated in front of a backdrop that includes a window with a view of a cityscape. The video is likely a news report discussing the marriage of Princess Eugenie, the youngest daughter of Sarah Ferguson and the Duke of York. The man is speaking, and his facial expression suggests that he is engaged in a conversation or delivering information. The style of the video is typical of a news broadcast, with a focus on delivering information in a clear and concise manner. The setting suggests that the video was filmed in a studio or newsroom."
+    ],
+    "video_ids": [
+        "tJvPcFLqtI4_7_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, seated in front of the backdrop.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a suit and tie, seated in front of a backdrop that includes studio elements like screens and windows, which aligns with the description. The man is clearly visible and the setting matches the specified condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed a suit and tie, seated in front of a backdrop that appears a elements like a and a, which ares with the description of The presence is dressed the and appears setting is the typical conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1039a84bf92b5d4cf84b2f9adf578992f7f30a87
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1262-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:925695d9019478fe9988dd42f4214e521b4bb592012ea5fe2e9b2fab4003505f
+size 144761
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..0668f449423ec3a545b3fbb1072cf287f4cd94c9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a gray t-shirt, who appears to be in a state of deep thought or prayer. He is standing against a backdrop of a dark blue sky filled with white and purple sparkles, giving the impression of a starry night. The man's hands are clasped together in front of him, suggesting a moment of reflection or meditation. The overall style of the video is calm and introspective, with the sparkly background adding a touch of whimsy to the scene. The man's position in the center of the frame, along with the sparkly background, draws the viewer's attention directly to him, emphasizing the importance of his actions and thoughts in this moment."
+    ],
+    "video_ids": [
+        "FzxLfkRFZMw_3_0to110"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a gray t-shirt with his hands clasped together.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a gray t-shirt with his hands clasped together in front of him. The background has a starry effect, but this does not contradict the core description. The man's posture and attire match the specified condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a gray t-shirt with his hands clasped together in a of him. The background is a starry,, but this does not contradict the core description of The man's attire and attire match the description '.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8ee0fdfc6e639bdf32bf5206a5683ae2b0db4658
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:673cba1c634b371051bbf46c6606031d38ab62aae563f1a0f3b12b43145f7bb4
+size 122295
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..e4293fd17d75592715070644b66fa99348aea913
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a car's engine compartment being worked on in a snowy environment. The first frame shows the engine compartment open, with a yellow device and a red cable connected to it. The second frame shows the addition of a black and white box on top of the engine. The third frame shows the engine compartment closed, with the yellow device and red cable still connected. The style of the video is a time-lapse, capturing the process of working on the car's engine in a cold, snowy setting."
+    ],
+    "video_ids": [
+        "mj1_ZfAVGvk_0_0to147"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Engine compartment, yellow device, red cable, black and white box? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the engine compartment of a Mercedes-Benz car with a yellow multimeter (yellow device), red cables connected to it, and a black and white box (likely a battery charger or similar device) placed on top of the engine. These objects are prominently visible and match the described elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows an engine compartment of a car-Benz,, a visible deviceimeter devicedevice device) a cables, to the, and a black and white box (likely a battery) or similar device) being on top of the engine. The elements match consistent displayed and match the description elements.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c6c7adc21808e0056a62f7e0856a3265907ac083
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d4db122534b1ec43f95c4e276d7e5c6993204428c5331923e4765c5a58690cb
+size 157330
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..76b468142ae48ff5bfbc51478f1601c8e77bd80f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a red and black sports car as it travels down a desert highway. The car's sleek design and vibrant colors are reflected in the car's curved windshield, which offers a clear view of the road ahead. The driver's seat is visible, with the steering wheel and dashboard in focus. The car's speed is suggested by the blurred scenery outside the window, which includes a yellow dividing line on the road and a clear blue sky above. The overall style of the video is dynamic and exciting, capturing the thrill of driving a high-performance sports car on an open road."
+    ],
+    "video_ids": [
+        "Q7XfzAV48J0_23_0to200"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red and black sports car, steering wheel, dashboard, road scenery, yellow dividing line, blue sky? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a red and black sports car with visible steering wheel and dashboard. The road scenery with a yellow dividing line and blue sky are also present, matching the description. No conflicting elements are observed.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a red and black sports car interior a elements wheel and dashboard. The road scenery is a yellow dividing line and blue sky is also present, indicating the description. The additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4a47a6e9daa7694796862175c1cbc4b4b2b86fd3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:39938c53f0c1d3874a4d7ddfe045a5235370cfd718040f15b773ca052a4f9243
+size 161944
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..63b6633d944747a4487530b302b5a87c569e067b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up of a tiger's face in a dense forest. The tiger, with its distinctive black stripes, is the main focus of the video. The camera is positioned in such a way that it captures the tiger's face in detail, from its eyes to its nose. The background is filled with lush green foliage, creating a natural and serene setting. The video is shot in a realistic style, with the camera remaining still, allowing the viewer to fully appreciate the beauty and majesty of the tiger in its natural habitat. The overall composition of the video emphasizes the tiger's presence in the forest, making it a captivating and engaging visual experience."
+    ],
+    "video_ids": [
+        "U1LhgusHlA8_79_49to226"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A tiger with distinctive black stripes.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a tiger with distinctive black stripes on its orange fur, which matches the description. The tiger's facial features, including its eyes, nose, whiskers, and striped pattern, are visible and consistent with the expected appearance of a tiger. There are no elements in the video that contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts depicts a tiger with distinctive black stripes, its fur fur. which is the description of The tiger's facial features, including its eyes, nose, anders, and the pattern, are prominently and consistent with the typical appearance of a tiger. The are no additional in the video that contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..339f67890c7d94016d8c52c2b945522248d7b62a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f08c0a45f64b9663b749dc256d09bacfb68e34c4f06174e811453c6f34c4aa0f
+size 233013
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..075177833e4216d4f3cc4829e2449901855f7522
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a weather forecast. The style of the video is a typical news broadcast with a professional appearance. The weatherman, dressed in a suit and tie, stands in front of a city skyline at night, with the lights reflecting off the water. He is holding a weather map and pointing to it, explaining the forecast. The forecast indicates a warm and breezy night with a few showers and wind speeds of 15-20 mph. The video is likely part of a local news program, as indicated by the logo of the news station in the corner. The overall tone of the video is informative and serious, with the weatherman presenting the forecast in a clear and concise manner."
+    ],
+    "video_ids": [
+        "5EGMYtNzsZ0_4_0to197"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Weatherman, weather map? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a weatherman presenting a weather forecast, with a visible weather map displaying temperature, conditions, and wind speed. The elements align with the 'Weatherman, weather map' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video features a personman holding a weather map, which a weather weather map in various and pressure, and other patterns. The presence in with the 'Weatherman, weather map' condition, contradict.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..492f6dde04001444d9d84f00f550306fab25433a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:962447fa4af133d41bb0b0245e6549b7dc2f8bb2887de831c838c757a9a5a762
+size 202685
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..6ae1cb7d891345c922cd7f32c1520aed6c7b2bd5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a professional, corporate-style video featuring two men in suits. The first man is standing with his hands in his pockets, looking at the second man who is standing with his hands on his hips, looking at the first man. The men are in a modern office setting with white walls and large windows. The video has a blue overlay with white text that reads \"9. YOUR FRIENDS WANT THE SIMPLE LIFE.\" The overall style of the video is sleek and professional, with a focus on the interaction between the two men."
+    ],
+    "video_ids": [
+        "KiOgWjClYGs_44_65to272"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men in suits.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men dressed in suits walking and talking in an office setting, which matches the 'Object(s): Two men in suits' condition. The additional elements, such as the text overlay and background, do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two men dressed in suits, side interacting in an office setting. which align the descriptionObject(s)' Two men in suits.' condition. The presence text such such as the text overlay and the, do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..faf7d5c7a9c3a5d3d321d0e562872e7264bc090a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a42bce54ab3dc75d2bc8dec491fecb3ef9cd6c0816cb28eeb26130a7377eb321
+size 89923
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..c05ec688f087c581c6631fd163f10c495786fdcf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man working on a black truck in a garage. The truck is parked on a lift, and the hood is open, revealing the engine. The man is wearing a blue shirt and is standing on the passenger side of the truck. He is using a tool to work on the engine. The garage is well-lit, and there are shelves in the background. The style of the video is a straightforward, real-life depiction of a mechanic working on a vehicle."
+    ],
+    "video_ids": [
+        "-12CyV1zcTc_17_0to199"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black truck, a man, a tool, the engine.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black truck with its hood open, a man working on it, and the engine compartment is visible. Although no specific tool is clearly shown being used, the man is interacting with the vehicle's components, which aligns with the 'Object(s)' condition. The presence of the engine is implied by the open hood and visible components like the radiator and headlight assembly.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black truck with its hood open, indicating man in on the, and various presence is is visible, There the specific tools is clearly visible in used by the man appears positioned with the engine in engine, which impliess with the 'tool(s)' condition. The presence of the garage and evident by the open hood and the engine. the engine and enginelight,.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a1b4da0730cd9c1616ee04b533532a860991c7c9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07219221804e27ebe02ede23a529edc2640f500e3c9c7d5cb8039ce04c6a2887
+size 130847
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..e154aa82961960ff42a489ec1e117b8964867efe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a meal being served on a plate. The plate is placed on a wooden table, and the meal consists of a piece of meat, a lemon wedge, and a side of fries. The meat appears to be cooked and seasoned, and the lemon wedge is placed on top of it. The fries are golden brown and crispy. The plate is round and has a blue color. The wooden table provides a warm and rustic background for the meal. The video captures the details of the meal and the setting, creating an appetizing and inviting scene."
+    ],
+    "video_ids": [
+        "uqiiUzj8uog_17_25to240"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A piece of cooked and seasoned meat, a lemon wedge, and a side of golden brown and crispy fries.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a plate with a piece of cooked and seasoned meat, a lemon wedge on top of the meat, and a side of golden brown and crispy fries. These elements match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a plate with a piece of cooked and seasoned meat, a lemon wedge, the of the meat, and a side of golden brown and crispy fries. The elements match the description provided, the questionObject(s)' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..64b74efac0caea88ea5720bc15172870c6feab7e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1263-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c53a325a88095ea39629687e312f7b9b01e6a11386508c3ad2d6a0ef06d910b
+size 160470
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..389e9f810e0cbe61b5a6a08ca7e1f743a9bd255f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a man in a suit and tie, who appears to be a news anchor or reporter. He is standing in front of a backdrop with the text \"THE DAMAGE REPORT\" and \"TRUMP TREMBLING\". The man is gesturing with his hands, possibly explaining or emphasizing a point. The style of the video is a typical news broadcast, with a focus on delivering information about political topics. The man's attire and the professional setting suggest a serious and formal tone for the segment."
+    ],
+    "video_ids": [
+        "OUZFap4EM6A_9_0to160"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man dressed in a suit and tie, which matches the 'Object(s)' condition. The man is the central focus, and his attire is consistent with the description. Additional elements such as the background and text overlays do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a man dressed in a suit and tie, which matches the descriptionObject(s)' condition. The man is standing central figure of and his attire is consistent with the description. The elements such as the background and the do do not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..acc7a01d738d90f6875f6dd947cc5673d6b8b580
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2fc59e0b2571b901387f6d514d20699304053d5a1b0dc4adb626ffe40cfec17c
+size 223857
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..977a499e95cbf26a856f9043451b8854c5196b3f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a small orange kitten exploring a lush green field. The kitten, with its fur glowing in the sunlight, is seen walking on all fours, its tail held high in the air. The kitten's curiosity is evident as it sniffs the ground, possibly investigating a scent or an object. The field is dotted with yellow flowers, adding a pop of color to the verdant landscape. The kitten's journey takes it through the grass, which sways gently in the breeze. The overall scene is one of tranquility and innocence, as the kitten navigates its way through the natural world."
+    ],
+    "video_ids": [
+        "rr7iRoBztNI_51_0to178"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small orange kitten? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small orange kitten walking through a grassy area with dandelions, which matches the description. The kitten's appearance and actions are consistent with the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a small orange kitten amidst through a fieldy area with yellowandelions. which align the description of The kitten's color, the are consistent with the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b29e3e796741ca7ac2ab885194882d6cc805de03
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3ac89f83c84f50fa5067507d95686f0cf4acc7b4fc97bb28d4d6f12c20392250
+size 263560
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..8e791fcb13a9efa2c3415b52c9b334877c96756f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up shot of a motorcycle tire, showcasing its tread pattern and design. The tire is black with a series of grooves and sipes that are typical for motorcycle tires, designed to provide grip and traction on various road surfaces. The motorcycle is parked in a garage or workshop, with a concrete floor and metal beams visible in the background. The lighting is bright, highlighting the details of the tire and the surrounding environment. The style of the video is straightforward and informative, focusing on the tire's features without any additional context or narrative."
+    ],
+    "video_ids": [
+        "4SXKDNAcgbQ_11_0to167"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A motorcycle tire with a black color, featuring a series of grooves and sipes.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a motorcycle tire that is black in color and clearly shows a detailed pattern of grooves and sipes on its tread. These elements are consistent with the description provided, and no conflicting visual elements are present that would contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a motorcycle tire with is black in color. displays displays a series pattern of grooves and sipes. its surface. The features are consistent with the description provided, indicating there additional elements elements are present in would detr the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9d2d59233ac68dc642fcc853ff895db4491a2bdf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0bfdf34fe297a12168a7b590059958d88346840d7261b010bff92f7dd6ec401
+size 62422
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..b65cd1edbe2b7eeb499024e3828e05e332298f74
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a football player in action, showcasing his athletic prowess and determination. The player, dressed in a white and blue uniform with the number 99, is seen in three distinct moments. In the first frame, he is seen in a defensive stance, his arms outstretched in a display of readiness. His eyes are focused, and his body language speaks of anticipation.  In the second frame, the player is seen in a dynamic pose, his body leaning forward as he prepares to make a move. His muscles are tensed, and his gaze is directed towards the field, indicating his readiness to spring into action.  The third frame captures the player in a moment of triumph. He is seen raising his arm in a victorious gesture, his face lit up with a smile. The background reveals a stadium filled with spectators, their faces a blur of excitement and anticipation. The player's triumphant pose suggests a successful play, adding to the overall narrative of the video.  Throughout the video, the player's uniform and the stadium in the background remain consistent, providing a sense of continuity. The video is a dynamic portrayal of a football player's journey, from anticipation to action to triumph."
+    ],
+    "video_ids": [
+        "-iIIpqbeSds_19_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A football player in a white and blue uniform with the number 99.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a white and blue uniform with the number 99 prominently displayed on the sleeve and chest. The player's attire, including the NFL logo and team name (Texans), matches the description. The background and lighting are consistent with a stadium setting, and there are no elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a football player wearing a white and blue uniform with the number 98. displayed on the back. the. The player is attire and including the helmet-style and the colors,likelyans), align the description. The player and setting are consistent with a professional setting, which the are no additional that contradict the core description.\"\n}<|im_end|>\nGuidId",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..828422b1fe118cb5baf4ad28305789c88dbbbaf5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:98c93a0cd6002383fbb3f0b6d52f8e5f6638cee0992b8201802c9c86ba37d627
+size 268931
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..8efda17a9b1b3643ecd1075c8059c5903ca45957
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a cake in a metal pan, with a focus on the center of the cake. The cake appears to be a golden brown color, and it has been cut into slices. The slices are arranged in a circular pattern around the center of the cake. The cake is placed on a wooden surface, and there is a small amount of liquid visible on the surface of the cake. The style of the video is a simple, straightforward shot of the cake, with no additional elements or actions taking place. The focus is solely on the cake and its presentation."
+    ],
+    "video_ids": [
+        "82Cbg5Dg9Kg_71_26to177"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Cake, metal pan, wooden surface, liquid? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a round, brown, textured object that resembles a cake, placed in a metal pan. The pan is resting on a wooden surface, and there is a golden-brown liquid (likely syrup or honey) pooling around the cake. These elements align with the described objects: cake, metal pan, wooden surface, and liquid.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dish, baked, cake object that resembles a cake, placed in a metal pan. The pan is resting on a wooden surface, and there is a liquid liquidrown liquid visiblelikely syrup or gl) surrounding around the cake pieces The elements collectively with the description objects: Cake, metal pan, wooden surface, and liquid.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..298716cd6ceb23ca770bf351ed7367ad6fc09756
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:08fe1bbd507a3f531c7a4196df99f3cc78f4c098ca93a4749a08045d3eb16ae0
+size 171984
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..8ffddf766f936d3c3fc0d575973e9f5021b399d5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated sequence featuring a purple cartoon bunny sitting on a brass horn. The bunny is positioned in the center of the frame, looking directly at the camera with a relaxed posture. The horn is large and shiny, with a golden stand that elevates it off the ground. The background is dark and moody, with a hint of a red curtain and a stringed instrument, possibly a harp, in the background. The lighting is soft and warm, highlighting the bunny's fur and the brass of the horn. The overall style of the video is whimsical and playful, with a focus on the cute and charming bunny character."
+    ],
+    "video_ids": [
+        "89_sTcGJvAQ_65_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A purple cartoon bunny, a brass horn, a golden stand, a red curtain, and possibly a harp.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a purple cartoon bunny sitting on a brass horn, which is placed on a golden stand. In the background, there is a red curtain, and a harp is partially visible behind it. All the specified objects are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a purple cartoon bunny sitting on a brass horn, which is placed on a golden stand. The the background, there is a red curtain, and the harp is partially visible, the. The the elements elements are present in match with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b982b73589a1569b6d5d12dd962ab4d993dbb8c7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b62a98b4ace195c6316ce2556e30f22c296f8a9afce0d771492629304f799583
+size 125439
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..b1cf3c643668e8e89737ed2e3621d92535d05dd3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a baseball player in the midst of a pitch. The player, dressed in a white uniform with red lettering, is seen from the back, his arm extended in the follow-through of the pitch. The player's focus is intense, his gaze directed towards the batter, ready to react to the swing. The background is a blur of spectators, their attention riveted on the unfolding action. The video is a dynamic snapshot of a moment in a baseball game, capturing the tension and anticipation of the pitch."
+    ],
+    "video_ids": [
+        "lGXkThQCCNg_4_23to159"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A baseball player in a white uniform with red lettering, a ball, and spectators.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a baseball player in a white uniform with red lettering (\\\"SOX\\\" and the number \\\"42\\\" are visible), holding a baseball, and spectators are blurred in the background, fulfilling the core description. The presence of additional elements like the logo and text overlay does not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a baseball player in a white uniform with red lettering,indicRX\\\" and the number \\\"10\\\"), visible on throwing a bat in and there in visible in the background. indicating the ' description of The player of the elements like the glove on the on on not contradict the main.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..eb862728bfbbfcf31c6851d3a14947af24b85bc6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd27de4207bf030315fe07c5e142fd64d331c53a390c0e4f04aa713580880907
+size 221157
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..fa56bded45f83ade69f21ad5b305bcefa10544ee
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a black cat with white paws and a white chest, intently observing a goldfish swimming in a small aquarium. The aquarium is filled with blue gravel and contains a few green aquatic plants. The cat's attention is fixed on the fish, which moves gracefully through the water. The setting appears to be indoors, possibly in a kitchen or dining area, as there are various household items visible in the background, including a white countertop, a wooden tray, and some bottles. The lighting is bright, suggesting it might be daytime. The cat remains stationary throughout the sequence, its gaze never leaving the fish. The camera angle is slightly elevated, providing a clear view of both the cat and the aquarium. There are no significant changes in the scene; the focus remains on the interaction between the cat and the fish."
+    ],
+    "video_ids": [
+        "03b13116f1418a4d0e804d99d13deb960761814f61de3f9aff9cc8ed247e5e7b"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Black cat with white paws and chest, goldfish in an aquarium filled with blue gravel and green aquatic plants.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black cat with white paws and chest interacting with an aquarium containing goldfish, blue gravel, and green aquatic plants. The core elements described in the condition are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a black cat with white paws and chest, with a aquarium. afish, blue gravel, and green aquatic plants. The elements elements of in the question are present and accurately depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..22181b50f92787443d1cee1b86358fc1ece2cc3c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1264-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d08a648718250ff5be644adb1fd78a348d7abd8c7fa530e1873cdab576b7d886
+size 274771
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..5da62070d68a56454bdbe4f62814f9f594b0cf20
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows an older man in a green shirt and white apron, cooking in a kitchen. He is using a spoon to taste the food he is preparing. The kitchen is well-equipped with various utensils and appliances, including a stove, oven, microwave, and refrigerator. The man appears to be focused on his task, and the kitchen is filled with the warm glow of the cooking process. The video captures the essence of home cooking, with the man's dedication to creating a delicious meal."
+    ],
+    "video_ids": [
+        "MuMGiK-1cnU_13_72to284"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An older man in a green shirt and white apron? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an older man wearing a green shirt and a white apron, which matches the description. He is in a kitchen, cooking, and the core subject is accurately represented despite other kitchen elements being present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an older man wearing a green shirt and a white apron, which matches the description. He is actively a kitchen setting cooking eggs which the scene elements of consistent represented.\"\n the elements elements being present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..274a8d23920e67f81f631a0b3a05fed4eddafcdb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ed085d9d4b95947ff4ed575f20a0f98ad554f405725ce85bac9556b70cba829
+size 181388
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..32e5e0113b852907942b7d957ce6a5986c1e78ec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a thrilling off-road adventure. A black pickup truck, equipped with a roof rack and a spare tire, is seen navigating a rocky trail. The truck's license plate is clearly visible. The driver, dressed in a green jacket, is seen in the first frame, skillfully maneuvering the vehicle. In the second frame, the truck is seen climbing a steep incline, its wheels digging into the loose gravel. The third frame shows the truck successfully reaching the top of the hill, its rear wheels suspended in the air. The backdrop of the scene is a rocky cliff, adding to the sense of adventure and challenge. The video is a testament to the truck's off-road capabilities and the driver's skill."
+    ],
+    "video_ids": [
+        "mUUHoQWpxVc_60_76to228"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Black pickup truck, rocky trail, rocky cliff, driver in a green jacket, license plate.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black pickup truck navigating a rocky trail with a rocky cliff on either side. A person in a green jacket is visible near the truck, and the license plate is clearly visible on the rear of the truck. All elements described in the condition are present and consistent with the video content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a black pickup truck driving a rocky trail. a rocky cliff in the side. The person wearing a green jacket is visible in the truck, and the truck plate is clearly displayed on the back of the vehicle. The elements of in the condition are present in match with the video content.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ab08682ba8fde646dda8803f45b1c1d0377e6b99
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:46fed8071ce48e7ee31fb197bac42d9f76071d35a7a8006d119d58a9e57af111
+size 258941
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..57713740e84a01ca2c65c4f51ae0bd2a81c78293
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with blonde hair, wearing a black and blue striped top, standing in a kitchen. She is speaking into a microphone, which is attached to her shirt. The kitchen is equipped with a white refrigerator and a microwave. The woman appears to be engaged in a conversation or recording a podcast. The style of the video is casual and informal, with a focus on the woman and her surroundings."
+    ],
+    "video_ids": [
+        "2gDAhogoeBc_13_0to154"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with blonde hair, wearing a black and blue striped top, and a microphone attached to her shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The woman in the video has blonde hair, is wearing a black and blue striped top, and has a microphone attached to her shirt. These elements are clearly visible and match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video in the video has blonde hair and is wearing a black and white striped top, and there a microphone attached to her shirt. The elements match consistent visible and match the description provided.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..63fd808d33726e02dbd0c7a9468913a955934fca
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a93ecda14be4e033dac47d5a0f9c00ad5cbcb5188626f62cf5be08d99bbb19b4
+size 217575
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..d644d87a12f31369bf8f65aaf678583014314b77
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are standing in front of a brick wall. The woman has pink hair and a tattoo on her arm, while the man is wearing a colorful hat. They are both looking at something off-camera, and the woman is holding her hair back. The video has a casual and relaxed vibe, with the two individuals appearing to be in a good mood. The brick wall in the background adds a rustic touch to the scene."
+    ],
+    "video_ids": [
+        "UvdBm9R0Jbg_8_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman standing in front of the brick wall; the woman has pink hair and a tattoo, the man wears a colorful hat; the woman is holding her hair back; both are looking off-camera.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a woman standing in front of a brick wall. The woman has pink hair and a visible tattoo on her arm, and she is holding her hair back with a clip. The man is wearing a colorful hat. Both individuals are looking off-camera, which matches the description. The video does not contradict any of these core elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a woman standing in front of a brick wall. The woman has pink hair and a tattoo tattoo on her arm, and the is holding her hair back. her hand. The man is wearing a colorful hat and Both individuals are looking off-camera, which align the description. The presence does not contradict the of the elements elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0ede8a37ddd3278320d2895d5918ba57a745effe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d6cf2c65bd1d9ec86238a34b88441851b8b6de513e4071ccddd6b6960b357aa9
+size 211763
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..d071086560040a542ba3571b71fe4d541e64f5da
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a white shirt sitting in front of a wooden structure, possibly a gazebo or a pergola. He appears to be engaged in a conversation or giving a presentation, as he is gesturing with his hands while speaking. The setting is outdoors, with a clear sky and greenery in the background. The style of the video is casual and informative, with a focus on the man and his surroundings. The overall atmosphere is relaxed and inviting, suggesting that the man might be discussing a topic related to the wooden structure or the outdoor setting."
+    ],
+    "video_ids": [
+        "4-ITBrINa2Q_28_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a white shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a white shirt with vertical stripes, which matches the description. The shirt is clearly visible throughout the frames, and there are no conflicting elements regarding the attire.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man wearing a white shirt, rolled stripes, sitting align the description of The setting is clearly visible, the frames, and the are no additional elements that the object.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..674be146c36ffeef4f78b7de9110c20c5459edc7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41e918c369f58909659e6800977191aeefd5648a744bf9dcdb03645c0198ece7
+size 167459
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..103b00230137723de8d4d9520b603a734a8e325a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a bustling city scene with a multi-lane highway filled with a variety of vehicles, including cars, buses, and trucks. The highway is surrounded by lush greenery and trees, providing a stark contrast to the urban environment. The vehicles are moving in both directions, indicating a busy time of day. The highway is lined with multiple traffic signs, including speed limit signs and directional signs, guiding the flow of traffic. The overall style of the video is a dynamic and colorful representation of urban life, capturing the energy and movement of the city."
+    ],
+    "video_ids": [
+        "32t8d2eQJAA_67_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Vehicles (cars, buses, and trucks), traffic signs (speed limit and directional).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows numerous vehicles including cars, buses, and trucks moving along a multi-lane highway. Traffic signs indicating speed limits (e.g., 60, 70, 80 km/h) and directional guidance (e.g., '\u5c0f\u578b\u8f66' for small vehicles, '\u5927\u578b\u8f66' for large vehicles) are prominently displayed. These elements fully satisfy the 'Object(s)' condition as described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a vehicles, cars, buses, and trucks, along a highway-lane highway. There signs, speed limits andnot.g., '40) 40) 80)/h) and directional information aree.g., straightExit\u8eca' which small vehicles) '\u5927\u578b\u8f66' for large vehicles) are also displayed. The elements align meet the 'Object(s)' condition as described.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0d9a6d1a49662901cbe7b4f25f4baa92d1ceebdc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d43c6604511fee2bf7854f81df8fae5c163e91dcd06e16c5cd4aab63a566a96
+size 317704
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..e82d2c389b19cc5624af21f8b351aac01effe33a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a blue and white plaid shirt, wearing glasses, sitting at a desk in an office environment. He is gesturing with his hands as he speaks, indicating that he is engaged in a conversation or explaining something. The office is well-lit and has a modern feel, with a large window in the background that lets in natural light. There are other people in the background, suggesting that the office is busy and active. The man appears to be the main subject of the video, and his actions and expressions are the focus of the scene. The overall style of the video is casual and informal, capturing a moment in the man's day at work."
+    ],
+    "video_ids": [
+        "KV05fU7ktMI_1_33to181"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue and white plaid shirt, wearing glasses, and several other people in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing glasses and a blue and white plaid shirt under a dark vest, which matches the description. In the background, other people are visible, though slightly out of focus, confirming the presence of several other individuals. The core description is accurately fulfilled without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a man wearing a and a blue and white plaid shirt, a blue blue, which align the description. The the background, there people are visible, though they out of focus, which the presence of additional other individuals. The setting elements is largely represented.\"\n any.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..50a6e453498e985f562a7ade12191d5e74e7e49f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3ac5e7408b30d003b0e987d2ed38d0acde5f384376d7c50d4ea78fe7bb97908
+size 184023
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..fdbd9020dbbe34979bf749dd8a85b3c689007970
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen grilling food on a barbecue grill. He is wearing a black jacket and is smiling as he cooks. The grill is placed on a sidewalk next to a brick wall. In the background, there is a parking sign indicating that it is a monthly parking area. The man is using tongs to flip the food on the grill. The video captures the man's enjoyment of grilling and the outdoor setting."
+    ],
+    "video_ids": [
+        "RBoObpZ3Ics_6_103to302"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a barbecue grill, tongs? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man standing next to a barbecue grill, using tongs to handle food on the grill. These core objects are accurately represented without contradiction, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man standing next to a barbecue grill. actively tongs to handle food. the grill. The elements elements ( present depicted, any.\"\n fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..897e91162de281b6d0f8880399018436d7e9cfda
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1265-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e8bc94c8a0c858681c89e0431e6cb2d0dd49fd4988cca001d5e93a2222c033c
+size 212037
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..83198d81e31fe08322517ba62b9b77d606c3f960
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a luxury sports car, focusing on the steering wheel and dashboard. The steering wheel is black with a silver emblem in the center, and the dashboard is sleek and modern, featuring a digital display. The car's interior is well-lit, highlighting the high-quality materials and craftsmanship. The video is shot from the perspective of the driver, providing a sense of the car's design and features. The style of the video is sleek and professional, with a focus on the car's interior and design elements."
+    ],
+    "video_ids": [
+        "XVkOpLoLfHc_38_0to102"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, dashboard? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a steering wheel and dashboard as the central focus, with the steering wheel prominently displayed in the foreground and the dashboard visible behind it. The details of both elements are consistent with the description, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a steering wheel and dashboard, described main elements. which the steering wheel prominently displaying in the foreground and the dashboard instruments in it. The presence such the the, consistent with the description, and there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..57d00f4630507192bd4f4e54fd4bfb5375b9649f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e6e0c868632ac1da7d7a72eb70dd0f0aa585f92cd265212a1aafca49a2c53d37
+size 209093
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..82551239c6a39fb90a3c8f348016aef026f94701
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are standing in front of a stone building. The man is wearing a green coat and a blue scarf, while the woman is dressed in a purple jacket and a white scarf. They are both looking off to the side, possibly at something or someone that has caught their attention. The building behind them is made of stone, adding a sense of age and history to the scene. The overall style of the video is realistic, capturing a moment in time between the two individuals."
+    ],
+    "video_ids": [
+        "4tHKBEmDo6g_2_0to135"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman standing together outdoors. The man is on the left, wearing a green coat and blue scarf, and the woman is on the right, wearing a maroon jacket with a fleece lining. Their presence and appearance match the 'Object(s): A man and a woman' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man and a woman standing side,. The man is wearing the left wearing wearing a green coat and a scarf, while the woman is on the right, wearing a purpleoon coat and a white scarf. Both attire and attire fulfill the descriptionObject(s)' A man and a woman' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..23666ad8ed7c53636543f8aa69df271b30ea9aa7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6d1444e9a87b071acc94900ac2cfdf3d58ac3e12ebfa8758a6be2f579f9b83a
+size 131490
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..3e440a32f34a031d933aaea951f549e198dabde6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a white car being worked on in a garage. The car is lifted on a lift, and the hood is open, revealing the engine and various wires and components. The garage is filled with tools and equipment, and there are shelves in the background. The car is being worked on by a mechanic, who is visible in the background. The style of the video is a time-lapse, showing the progress of the car being worked on. The video captures the intricate details of the car's engine and the mechanics at work."
+    ],
+    "video_ids": [
+        "-dEd6uQjSvY_57_62to259"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white car, a mechanic, various tools and equipment, shelves? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a white car with its interior components exposed, including wiring and mechanical parts. The background includes shelves with various tools and equipment, consistent with a workshop or garage setting. Although no mechanic is visible, the presence of tools and equipment on shelves fulfills the described condition. The core elements are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a white car with its hood visible visible, indicating various and mechanical parts, The setting includes shelves with various tools and equipment, which with a mechanic setting garage setting. There there mechanic is visible in the presence of tools and the suggests the supportsfills the ' condition.\"\n The video elements of present represented.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fa8e0e10027665e08e19b0e4865186fb9cef362e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6449691a6f92623947ff2bb1cee4968e481730d6dd71db4772769a2964d93393
+size 267808
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..5adfed6bbc0ef85f492c4a2931f557d3ea20b103
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene backyard scene on a sunny day. A lush green tree dominates the left side of the frame, while a wooden fence runs along the right, separating the garden from neighboring properties. In the background, a house with a sloped roof is visible under a clear blue sky. The garden is vibrant with various plants and trees, some of which are in full bloom with white flowers.  A pigeon, with iridescent feathers displaying shades of blue and purple, gracefully flies into the frame from the right. It approaches a hand extended towards it, offering food. The bird lands lightly on the hand, its wings slightly spread as it takes a moment to settle. The pigeon then begins to peck at the offered food, its head moving rhythmically as it eats. The camera remains steady throughout, focusing on the interaction between the pigeon and the hand, capturing the peaceful and intimate moment amidst the tranquil garden setting."
+    ],
+    "video_ids": [
+        "40fb67d6df567fb90842637acd11c208a1a59071bad9598636c7e6b20d14da51"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Lush green tree, wooden fence, sloped-roofed house, various plants and trees, pigeon, hand.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a lush green tree on the left, a wooden fence running along the right side, and various other plants and trees in the background. A sloped-roofed house is visible in the distance. The central focus is a pigeon flying toward and landing on a human hand, which matches the described 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a lush green tree, the left, a wooden fence in along the background side, a a plants plants and trees in the background. A sloped-roofed house is visible in the upper. The presence focus is on pigeon being towards a being on a hand hand, which is the description elementsp(s)'.\"\n.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..614b8b4ccd6477f8686255cde76490e2ceeeabff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e8d95104865831c32c2c670a8176e769c5604b12570be8ca12e6d5576aa32c9a
+size 272771
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..8c08d5f2d9322accb6ac5c116efa5e169374f957
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a knight on horseback, riding through a forested area. The knight is clad in armor and holds a lance, suggesting a scene from a medieval or fantasy setting. The horse is also armored, and both the knight and the horse are in motion, indicating that they are in the midst of a journey or battle. The forest around them is lush and green, with trees and foliage filling the background. The lighting in the scene is natural, suggesting that the video was taken during the day. The overall style of the video is realistic, with attention to detail in the armor and the environment."
+    ],
+    "video_ids": [
+        "D98YwaKrjwQ_13_0to133"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A knight on horseback, both clad in armor and holding a lance.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a knight on horseback, both clad in armor, and the knight is holding a lance. The scene is consistent with the description, even though there are additional elements like the statue base, surrounding landscape, and architectural structures, which do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a figure on horseback, both clad in armor and and holding knight is holding a lance. The knight is set with the description provided and though the are no elements like the lush-like and which green, and the elements in which do not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..902a91d7b2f851e44c65a2a7328b2513c177384c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:94ff3461236d573abee9bf808bf7154dda782abce6ee547f92c56a6a7f44f850
+size 332600
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..6bd6d1e7919a8f088fc8d7588687a1065fb052d1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young man in a kitchen, wearing a white apron with a logo on it. He is standing in front of a black refrigerator, which has a silver handle and a control panel. The kitchen has a gray wall and a red bowl on a counter. The man is looking to the side with a thoughtful expression. The style of the video is casual and informal, with a focus on the man and his surroundings. The lighting is soft and natural, suggesting an indoor setting. The overall mood of the video is relaxed and comfortable."
+    ],
+    "video_ids": [
+        "-vafAXEGiKY_211_603to819"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Young man in a white apron, black refrigerator with a silver handle and control panel, red bowl.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man wearing a white apron, standing in front of a black refrigerator with a silver handle and control panel. A red bowl is also visible on the counter to the left. All elements described in the condition are present, and no conflicting elements are observed.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man wearing a white apron, standing in front of a black refrigerator with a silver handle and control panel. There red bowl is also visible on the counter. the left of The elements in in the condition are present in and there additional elements are introduced.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..39175eb0b5514ddae7b8926d76f1f0720ce1eec6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dc8070b9b1d6c1854fc7dbbca5d6496fbd4c1c79c22a4be4f8f5e63d5347f610
+size 125925
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..c80ae2e6c9c1c975bbb1bbae28f3dc1244b84004
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up portrait of a man with a serious expression. He has dark hair and a light complexion. He is wearing a black turtleneck and has a small earring in his left ear. The background is a plain white wall, which provides a stark contrast to his dark clothing. The lighting in the video is soft and even, highlighting the man's features without creating harsh shadows. The overall style of the video is minimalist and focused on the man's expression and the details of his appearance."
+    ],
+    "video_ids": [
+        "QsMoWjJmRKc_9_0to168"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man with dark hair, light complexion, black turtleneck, and a small earring in his left ear? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with dark hair, light complexion, wearing a black turtleneck, and a small earring in his left ear. These features match the description provided in the 'Object(s)' condition. The presence of a nose piercing and a patterned jacket does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with dark hair, light complexion, and a black turtleneck, and a small earring in his left ear. The elements match the description provided, the questionObject(s)' condition.\"\n There video of a plain piercing on the slighted background in not contradict the description description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c9d397917744eea102548dbecfb436c2f42047e6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a83a232ae076ee7caf4984d86712d099871c6cb2749d45e65c81fd669476aa8
+size 82518
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..219f7bdc9a5506e11b3895d577a0f1c8c7855bd4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a breathtaking aerial view of a mountainous landscape. The scene is dominated by a dense forest of trees, their green foliage contrasting beautifully with the rocky terrain. The mountains, blanketed in a layer of snow, rise majestically in the background, their peaks reaching towards the clear blue sky. The perspective of the video is from a high vantage point, providing a panoramic view of the landscape. The colors in the video are vibrant, with the green of the trees, the white of the snow, and the blue of the sky creating a stunning visual spectacle. The overall style of the video is realistic, capturing the natural beauty of the landscape with precision and detail."
+    ],
+    "video_ids": [
+        "UVLFhC-3slY_3_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dense forest of trees, rocky terrain, snow-covered mountains, clear blue sky? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by showcasing a dense forest of trees, rocky terrain (visible in the foreground and on mountain slopes), snow-covered mountains in the background, and a clear blue sky. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition by depicting a dense forest of trees, rocky terrain,visible in the form), mid the slopes), snow-covered mountains in the background, and a clear blue sky. The these elements of in present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1d0ea7e64bb93a918175ca1845a4976149570833
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1266-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:536890184c320b7b2a6ef55b29c4125c431b5518c0ec07d3c98e87e43e853a64
+size 293550
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..24db9d88a6c118635b904b831d8c5c9e5252da83
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a steering wheel in a car, with the camera moving from the left side of the wheel to the right side. The steering wheel is black and has a red and white logo in the center. The car's dashboard is visible in the background, with various controls and gauges. The steering wheel is made of leather and has a shiny appearance. The car's interior is red, and the steering wheel is positioned in the center of the dashboard. The camera moves smoothly, providing a clear view of the steering wheel and its surroundings. The video is shot in a realistic style, with no filters or special effects."
+    ],
+    "video_ids": [
+        "S5S_iAvnD2g_14_40to203"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, car's dashboard, controls and gauges? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a steering wheel with a Corvette logo, part of the car's dashboard, and various controls and gauges. The elements described in the condition are all present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a steering wheel with a logo logo, a of a car's dashboard with and various controls and gauges. The presence in in the question are present present and accurately depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..56b0efa61c3e822ef5764ac34711e5567b086c5b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8a3a372499d81fe3b60860f432b109867a4871ee4db68f692efc084675244108
+size 252224
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..ae26fc94df1d5dfa99d1795b8d8b0b9615a8f374
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with dark hair and a plaid shirt, who appears to be in a room with other people. She is looking directly at the camera with a surprised expression on her face. The room has a television mounted on the wall, and there are paintings hanging on the wall. The lighting in the room is bright, and the overall style of the video is casual and candid."
+    ],
+    "video_ids": [
+        "VF_S5wzJTeM_17_0to199"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with dark hair and a plaid shirt, and other people in the room.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman with dark hair wearing a plaid shirt, which matches the core description. Other people are also visible in the background, though blurred, which satisfies the condition of 'other people in the room.' There are no elements that contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a woman with dark hair wearing a plaid shirt. which ful the ' description. Additionally people are also visible in the room, fulfilling they, which ful the condition of 'other people in the room'. The are no elements in contradict the given.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..59816989a38e5940e7c4c17e21b9c5bd7e515791
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:774bbb3f115dfc3c16cac77dc3045b82ceff5584bebeb56a8ec4487bf78509dd
+size 197165
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..9ea3c6c432edfeac39e731e87f20b9abfbd216c1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman performing on stage. She is dressed in a white, lacy bodysuit with a corset-style waistband. The bodysuit is paired with high-waisted, beige boots that reach up to her thighs. The woman stands confidently with her hands on her hips, and she is holding a microphone in her right hand. The stage has a vibrant yellow background with red and white stripes. The lighting is bright, highlighting the woman and her outfit. The overall style of the video suggests a live performance, possibly a concert or a music video."
+    ],
+    "video_ids": [
+        "PlRrvJMuCYM_29_0to128"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a white, lacy bodysuit with a corset-style waistband and high-waisted beige boots; she is holding a microphone in her right hand.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a white, lacy bodysuit with a corset-style waistband and high-waisted beige boots, holding a microphone in her right hand. These details align with the description provided, and there are no conflicting elements in the visual.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a woman wearing a white, lacy bodysuit with a corset-style waistband and high-waisted beige boots. which a microphone in her right hand. The elements match with the description provided, making there are no additional elements in the video content}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2f1029561eb6a7d9990d18ec7e2d12d5cdd40213
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac3fe693914a497598e188ebf154638918521958e6281bf9ad107f90f926dbaf
+size 149800
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..4ee2ae893a3560ac8db94da8cbcf324b36d630f7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man with a beard and sunglasses, smiling and posing for the camera. He is wearing a black jacket with a white logo on the left chest. The man is standing on a bridge overlooking a body of water, possibly a river or a lake. In the background, there are other people, some of whom are also wearing sunglasses. The man is holding a camera, suggesting that he is taking a selfie. The overall style of the video is casual and relaxed, with a focus on the man's enjoyment of the outdoor setting."
+    ],
+    "video_ids": [
+        "Rqctpam4o0o_53_35to158"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and sunglasses, a black jacket with a white logo, other people, a camera? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and sunglasses, wearing a black jacket with a visible white logo (HYDE). Other people are present in the background, and the man is holding a camera (implied by the selfie-style video). All elements match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and sunglasses wearing wearing a black jacket with a white white logo.likelyD). There people are present in the background, and the man is holding a camera,implied by the selfie-style video). The elements match the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..73ea8cdc3b2bd01c348de0e74c27855cfcd3095e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37ef812b44408ec38f993007c9e61585ede19268773296604152f5775f36a69b
+size 193940
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..bfcb4a504836b15b53592301e6e5bf1e140b7284
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a cooking show featuring a chef in a kitchen. The chef is wearing a white chef's coat and is preparing food on a counter. The kitchen is well-equipped with various cooking utensils and ingredients. There are bottles of oil and spices on the counter, and a refrigerator in the background. The chef is using a knife to cut food and is also using a pan to cook. The kitchen has a window that offers a view of the city. The chef is focused on his task and appears to be in the middle of cooking a meal. The overall style of the video is informative and instructional, with the chef demonstrating cooking techniques and providing viewers with a glimpse into the culinary process."
+    ],
+    "video_ids": [
+        "tQtuV-LO_q4_11_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A chef in a white chef's coat, a knife, a pan, bottles of oil and spices, a refrigerator, and a window with a city view.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a chef in a white chef's coat, a pan on the stove, bottles of oil and spices on the counter, a refrigerator in the background, and a large window with a city view. These elements are all present and match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a chef in a white chef's coat, a knife, a stove, a of oil and spices on the counter, a refrigerator in the background, and a window window with a city view. The elements match all present and match the description provided.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..09c5a0cd33eace34a00549e8e7340614eebfb8e8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:439e2a2de40d05878bb0a667630796095b4455a6bdcf257744f8d367d88f1f70
+size 131700
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..1b183069dc7fae5e70cac6ec461e690802447e2f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young man and woman are seen in a room with a festive atmosphere. The man is wearing a colorful, multi-colored wig and a black bandana, while the woman is dressed in a red jacket and sunglasses. They are both seated in front of a table, with the man pointing at something on the table. The room is decorated with string lights, adding to the festive ambiance. The video captures a moment of fun and celebration between the two individuals."
+    ],
+    "video_ids": [
+        "YxhKak5OmP8_65_0to146"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man, a young woman, a table, string lights.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a young man and a young woman seated at a table, with string lights visible in the background. All core elements mentioned in the condition are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a young man and a young woman sitting at a table. engaging string lights in in the background. The the elements of in the condition are present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f26e8ccef8ac9b2afbdcee37c19d267f7aa53c6b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba25d9bb446f1ce382d99174e12ac73451674ccaf51c543ab4cfdc4328148a03
+size 225773
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..bbb5ed6957f4b93f509ec8128ed57aeb56cd141c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a kitchen, holding a chocolate-covered popsicle. She is wearing a red and white polka dot dress and has a blue ribbon in her hair. The kitchen is well-equipped with a sink, oven, and microwave. On the counter, there are several chocolate-covered popsicles, and a basket of apples is also visible. The woman seems to be pointing at the popsicles, possibly indicating her creation or the process of making them. The overall style of the video is casual and homey, with a focus on the woman and her culinary creation."
+    ],
+    "video_ids": [
+        "HxdusLakrAM_43_37to217"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a chocolate-covered popsicle, several chocolate-covered popsicles, a basket of apples? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman holding a chocolate-covered popsicle, with several other chocolate-covered popsicles on a wooden board in front of her. A basket of apples is also visible on the counter near the stove. All specified objects are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a woman holding a chocolate-covered popsicle, and several other chocolate-covered popsicles on a tray board in front of her. There basket of apples is also visible on the counter. the pops. The the elements are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..740302ca5d05414a9b1b44374cc5c49ced974133
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b7d4973a09d2d89a13d971dc6298e6351e97dd34f0355d77d859747b4c80e58
+size 126969
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..fdf67243e20eb8b69eb9f256e9704fefa670d0d7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman and a young girl are seen in a kitchen, engaging in a daily activity. The woman, dressed in a blue shirt, is standing at the sink, washing dishes. The young girl, wearing a green dress, is standing on a stool next to the woman, observing and learning. The kitchen is well-equipped with wooden cabinets, a sink, and various kitchen items such as a kettle, a bowl, and a bottle. The scene captures a moment of bonding and learning between the woman and the child, set in a warm and inviting kitchen environment."
+    ],
+    "video_ids": [
+        "I62TeuSuMPg_1_0to147"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a blue shirt and a young girl in a green dress.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a blue shirt and a young girl wearing a green outfit (appears to be a green dress or long-sleeved top with polka dots) standing at a kitchen sink. The core description is accurately represented, and no elements contradict this.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a blue shirt and a young girl in a green dress.aaring to be a dress dress). a skirtleeved top). aka dots). standing in a kitchen counter. The woman elements of accurately represented in with there additional contradict the.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d099182267f48807fed5a692de11511a749ce755
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1267-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2266be394979a9c960d828e376b05a4a08f99f9dbda77de9233e43b79a3bae17
+size 123654
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..57886c80cf71210addb6198cdc1fc2f13f5879ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up of a black sports car, specifically a Jaguar, with a focus on the front grille and license plate. The car is parked in a wooded area with trees and green foliage in the background. The license plate reads \"ONTARIO\", indicating the car is registered in the Canadian province of Ontario. The style of the video is realistic and it appears to be a still image rather than a moving one. The car is the main subject of the video, with the surrounding environment serving as a backdrop. The lighting in the video is natural, suggesting it was taken during the day. The car's design and the license plate are clearly visible, providing a sense of location and context."
+    ],
+    "video_ids": [
+        "da0mR3HYWe8_31_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Black Jaguar sports car (focus on front grille and license plate)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video focuses on the front grille and license plate of a black Jaguar sports car, as described. The camera pans across the front, clearly showing the Jaguar emblem, the grille, and the Ontario license plate. The car's black color and sports car design are evident, and no elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video focuses on the front grille and license plate of a black Jaguar sports car, which described. The car angle slightly the front of highlighting showing the grille emblem and the grille, and the license license plate. The background is sleek color and the car design are also, and the additional contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a762ac21c3bf4011e0e015af60d3505b097fbe5d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f41f69597b350518c6d123b101c28192e70a9d580ea5264c086bb0ba81217bd5
+size 135282
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..a1a5632669fcd93579b353de64b8fa2e8be1d6ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a bald head and a beard, sitting on a couch with a floral pattern. He is wearing a gray t-shirt and appears to be in a relaxed, indoor setting. The man's expression is neutral, and he is looking directly at the camera. The lighting in the room is soft and even, suggesting an indoor environment with natural light. The style of the video is straightforward and documentary-like, with no additional elements or effects. The focus is solely on the man, and there are no other objects or people in the frame. The video seems to be a simple, unadorned portrait of the man, possibly for an interview or a personal profile."
+    ],
+    "video_ids": [
+        "2rNWR2-9b58_62_101to319"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a bald head and a beard, wearing a gray t-shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a bald head and a beard, wearing a gray t-shirt, which matches the description. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a bald head and a beard, wearing a gray t-shirt. which matches the description provided The background includes additional elements in not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4f1560236393758af59663903cc4c765507c0305
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e147e3b2c408a9193d6a6c11e54578734d8776385aa79771a9709bfdf356cfdb
+size 99172
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..c22eb52f3bcb2103648929f403ae02ab99e5db80
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic scene at a race track. The main focus is a bright orange sports car, which is seen driving down the track in three different positions. The car is sleek and modern, with a low profile and large wheels. It's moving at a high speed, leaving a trail of dust behind it.  In the background, there are other cars parked on the side of the track. These cars are also sports cars, but they are not as prominent as the orange car. They are parked in a row, facing the same direction as the orange car.  The track itself is wide and well-maintained, with clear markings and a smooth surface. It's surrounded by a fence, which separates the track from the surrounding area.  The style of the video is action-packed and fast-paced, with a focus on the orange car and its movement. The camera follows the car as it drives down the track, capturing its speed and agility. The other cars in the background add depth to the scene, but they are not the main focus. The overall impression is of a high-speed race, with the orange car leading the way."
+    ],
+    "video_ids": [
+        "dcWCvKZUNw4_15_0to129"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: - Bright orange sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a bright orange sports car, specifically a McLaren, which is the central focus of the footage. The car is clearly visible and matches the description provided. Other cars and people are present in the background, but they do not contradict the core description of the bright orange sports car.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a bright orange sports car, which a Lamb, which is the main object of the scene. The car is shown visible and matches the description of. The elements are the in present in the background, but they do not detr the main description of the bright orange sports car.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5dbb0c67d2b1063935ef9bcf09ea8952560798a6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b9f95ffd0ff6186d5f08d4a87a6d16279322b8a6bea6678eeff8e0a6d0500d7c
+size 194338
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..a51226af510f0b7ff23e5f3591fde161bbb5a59a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a stylized, animated depiction of a group of characters in a desert-like environment. The characters are dressed in colorful, futuristic outfits and are armed with various weapons. They are standing on a large, golden chest, which is adorned with a pearl necklace. The characters are in dynamic poses, suggesting movement and action. The background features a clear sky and a mountain range, adding to the adventurous atmosphere of the scene. The overall style of the video is vibrant and energetic, with a focus on action and adventure."
+    ],
+    "video_ids": [
+        "Xj1GdfG-iGw_17_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Group of characters in colorful, futuristic outfits, armed with various weapons.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a group of characters dressed in colorful, futuristic outfits and armed with various weapons, such as guns and bows. The characters are clearly visible and match the description, with no elements contradicting the core requirement.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a group of characters dressed in colorful, futuristic outfits, equipped with various weapons, which as guns and energy. The characters are standing depicted and the the description provided and no additional thating the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5afbac168452648ee18d61e48761704944458c84
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:959039adae5a00c55ac96cd8f8cc8ac6b69ca4feff91227e39b6f0bf9a6cfa2d
+size 244564
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..3c384927c3c4cabfd6996ed70d8cf4f16f6881fd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit and tie standing in front of a blue background with a logo. He appears to be speaking or presenting, as suggested by his open mouth and attentive expression. The logo on the background is likely associated with the content or the channel the video is from. The man is dressed professionally, wearing a dark suit and a red tie, which adds a pop of color to his outfit. The overall style of the video suggests it could be a news segment or a professional interview. The lighting is bright and even, highlighting the man and the logo in the background. The video seems to be well-produced, with a clear focus on the man and the message he is conveying."
+    ],
+    "video_ids": [
+        "GslOsntnOyo_0_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a dark suit and red tie, blue background, logo? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed in a dark suit and red tie, standing in front of a blue background with a visible logo ('BALANCO GERAL' and 'AO VIVO'). All elements described in the condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man wearing in a dark suit and a tie, standing against front of a blue background with a logo logo.VIZA').ERAL'). and aB VIVO'). The elements of in the condition are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..16741aa3a52f42310a69b40f818fd99298eeebeb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bcee9556e52a92f5c0336fb7b2114d7da73a841bac6f5967ec142807f425c438
+size 138565
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..e1932d075af1de9d7723292e460d9b71fed3b05c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the breathtaking view of a waterfall cascading down a cliff into a river. The waterfall, with its powerful flow, is the central focus of the video. The river, filled with white foam, flows through a lush green forest, adding a sense of tranquility to the scene. The forest, dense with trees and foliage, provides a natural backdrop to the waterfall. The video is taken from a high vantage point, offering a panoramic view of the waterfall and the surrounding landscape. The overall style of the video is a mix of nature and adventure, showcasing the beauty of the waterfall and the ruggedness of the landscape."
+    ],
+    "video_ids": [
+        "Jgegx5BH5xM_4_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Waterfall, river, forest? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a waterfall cascading into a river, surrounded by a lush forest. These elements are prominently featured and align perfectly with the specified conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a waterfall cascading down a river, surrounded by a lush forest. The elements are all featured and align with with the ' '.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..31d800a1ad998deefb44d40b7fc08acbe5504672
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7458e7de9727182a9f753d57711e38d912c25e92175814aaa9f371f718d4b24f
+size 149291
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..9091e0d5bab9596b129b7b81fade66e3412ee4ec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a young woman sitting in a room, engaged in a conversation with another person whose face is not visible in the frame. The woman is wearing a black and white patterned sweater and has her hair pulled back. She is looking directly at the other person, suggesting a serious or intense discussion. The room appears to be a living room or a similar indoor space, with a window and a bookshelf visible in the background. The lighting in the room is soft and natural, indicating that the video was likely taken during the day. The style of the video is realistic and naturalistic, with no visible special effects or artistic manipulations."
+    ],
+    "video_ids": [
+        "NDSxhgwiGQQ_40_0to181"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman and another person (face not visible).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman facing another person whose face is not visible, matching the 'Object(s)' condition. The young woman is clearly visible, and the other person is partially visible from behind, with their face obscured, fulfilling the description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman wearing another person whose face is not visible. which the descriptionObject(s)' condition described The presence woman is wearing the, and the other person is partially visible, the, fulfilling their face not. which the requirement.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d4fbceeae4e5abfcb5c524bc54679ad2689c2e68
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63430d2b8477358ead6d48ece66aeab27fa9457484835eb9e29790a5af7e90de
+size 190788
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..56ccb347131504c14273b1d749a8aca466dc0658
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a cooking show featuring three hosts standing behind a white counter. The counter is adorned with various food items and cooking utensils. The hosts are engaged in a lively conversation, with one of them holding a book. The setting is a studio with a blue background, and the overall style of the video is professional and informative. The hosts are dressed in casual attire, and the lighting is bright, highlighting the food and the hosts. The video is likely aimed at viewers interested in cooking and food."
+    ],
+    "video_ids": [
+        "s2I0i5pMLq0_1_40to261"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three hosts, various food items, cooking utensils, and a book.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three hosts (a chef, a woman, and a man in a suit) standing behind a counter with various food items (burgers, skewers, etc.), cooking utensils (a pan, bottles), and a book (a tablet or book-like object held by the man). All elements described in the condition are present, and no conflicting elements are visible.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three individuals standingtwo man and a woman, and a man) a white) standing behind a counter with plates food items andplatesers, fruitsers, and.) cooking utensils (p frying, a, and a book (a red or a-like object). by the chef in The elements mentioned in the condition are present, and the additional elements are introduced.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a2f45b4e5a20ebdfbb8831ab728d959284fdf438
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1268-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1e3e77a066a633105f893dd7b3985889ff306bdf36016992c0455b64fed7fb9
+size 159709
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..0047bbb5e17245323082c5b527e6ad11c80f097c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a black BMW motorcycle on display at a motorcycle show. The motorcycle is positioned on a rotating platform, allowing viewers to see it from all angles. The motorcycle is sleek and shiny, reflecting the lights in the showroom. The showroom itself is spacious and well-lit, with other motorcycles on display in the background. The video captures the elegance and sophistication of the BMW motorcycle, highlighting its design and features. The rotating platform adds a dynamic element to the video, making it more engaging and visually appealing. The showroom setting provides a context for the motorcycle, emphasizing its status as a luxury vehicle. Overall, the video is a promotional piece for the BMW motorcycle, designed to showcase its design and appeal to potential buyers."
+    ],
+    "video_ids": [
+        "ClxqSTfNqpg_4_0to147"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Black BMW motorcycle? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a black BMW motorcycle on display, which matches the core description. The motorcycle is clearly identifiable as a BMW by its logo and styling, and it remains the central focus throughout the video. Additional elements like other motorcycles and exhibition signage are present but do not contradict the primary object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a black motorcycle motorcycle, a in which align the description description of The motorcycle is the visible as a BMW model the branding and design, and it is the central focus throughout the frames. The elements such the motorcycles in the-like in present but do not contradict the main focus description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0d58d21e3776ab3c920764b56a2130426e991a6f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ff647265766c22f1ebc930436e33dafd5bf5ecde2aa5558df74a3b51ede40b6
+size 180362
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..2f2e9a0f373bddc76dc9eb51afb7d9079e034dcf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a blue plaid jacket and a gold tie, smiling and gesturing with his right hand. The man appears to be in a good mood, possibly engaging in a conversation or presenting something. The background is a solid blue, which contrasts with the man's attire and highlights him as the main subject of the video. The style of the video is straightforward and seems to be a casual interview or presentation, with the focus on the man's facial expressions and body language."
+    ],
+    "video_ids": [
+        "5sCFl-VUyew_0_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue plaid jacket and a gold tie, smiling and gesturing with his right hand.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue plaid jacket and a gold tie, smiling, and gesturing with his right hand. These elements align with the description provided. The background and watermark do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a man wearing a blue jacketaid jacket and a gold tie, smiling, and gesturing with his right hand. The elements match with the description provided, The man is the are not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5f014bb06f8be6b61a4472a540eaa33ed9031ce6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1935b12a89f9500a0a81782c5abcc111993686927684dfd83f2298d3456e214a
+size 226782
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..ed262ca6bcb4d1b55f2cbdb42142efeedb0cca9e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the rear view of an orange sports car in motion. The car's design is sleek and aerodynamic, with a prominent rear wing and a large rear window. The car's taillights are bright and red, and the license plate is clearly visible. The car is driving on a road with a sidewalk and a building in the background. The car's speed and the blur of the background suggest that the car is moving quickly. The overall style of the video is dynamic and action-packed, capturing the thrill of driving a high-performance sports car."
+    ],
+    "video_ids": [
+        "-zAn220OaJA_12_480to689"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An orange sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features an orange sports car, specifically a Lamborghini Gallardo, which matches the description. The car is shown from various angles, highlighting its distinctive design and vibrant orange color. While there are other elements in the video (such as a person and a building), they do not contradict the core description of an orange sports car.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features an orange sports car, which focusing rearorghini Hurardo, as is the description of The car is shown from the angles, emphasizing its design features features color color color. The the are other elements in the background,such as the city and a city), they do not contradict the core description of the orange sports car.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4ba2ae836b948aceb358646337641c421317d447
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3bba22a0a46b34c414fa20d2f297c40fc2486403d3903da583d75c408ac43e09
+size 156619
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..5d6cf62f3947a52ea2d874684b3418c527cb79b4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a kitchen, standing in front of a counter. She is wearing a gray shirt and has her hands outstretched, as if she is presenting or demonstrating something. The kitchen is well-equipped with various appliances and utensils, including a refrigerator, an oven, a sink, and a cutting board. There are also several bottles, cups, and bowls on the counter. The woman appears to be in the middle of a cooking or baking process, as suggested by the presence of these items. The overall style of the video is casual and informative, likely aimed at teaching viewers how to prepare a particular dish or demonstrating a cooking technique."
+    ],
+    "video_ids": [
+        "HYQV8NdpWFc_17_0to130"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a gray shirt with her hands outstretched, several bottles, cups, and bowls on the counter.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a gray shirt with her hands outstretched, as described. Additionally, there are multiple bottles, cups, and bowls visible on the kitchen counter, fulfilling the condition. The presence of other kitchen elements does not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a gray shirt with her hands outstretched, standing described. There, there are several bottles, cups, and bowls on on the counter counter, which the ' of The presence of these kitchen items like not contradict the core.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7ab6fb1c2b93a72b93268e89c05a75eb035226a5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e61f561f4dca259555527d7886a44a5db4713228e918b995d3a145b023538c16
+size 89337
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..d4c4f45929e5e4b80ce94a1beceaae5c5c3a680d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car, focusing on the dashboard and steering wheel. The dashboard features a large touch screen display, which is turned on and displaying a map or navigation system. The steering wheel is black and has a logo on the center. The car's interior is predominantly white, with black accents on the dashboard and steering wheel. The style of the video is a straightforward, static shot of the car's interior, with no movement or action. The focus is on the technology and design of the car's dashboard and steering wheel."
+    ],
+    "video_ids": [
+        "MRQAWas3KXw_26_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dashboard with a large touch screen display, steering wheel with a logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a dashboard with a large touch screen display, which is centrally located and displaying a map interface. Additionally, a steering wheel with a visible logo (Toyota) is present on the right side of the frame. These core elements are accurately depicted, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a dashboard with a large touch screen display and which is a located and occupies a map.. The, there steering wheel with a logo logo islikely) is prominently in the left side of the dashboard. The elements elements match consistent depicted, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ae7c1ccd63123ff61ac4afdf219cc9ebcb55daf2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:505c2fde58e9aae840a52da997b0255f13f3f3b4e6f23382ef7e697757db4f74
+size 119345
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..2c55b87b71b1c5cb3bbca245a6c324da4d763418
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a blue shirt standing against a blue background. He is bald and wearing glasses. To his right, there is a can of Coca-Cola. The man appears to be speaking or explaining something. The text on the screen reads, \"If you put a can of Coca-Cola into a pool of water, it will sink.\" This suggests that the man might be discussing the buoyancy of the Coca-Cola can or perhaps the density of the liquid inside the can. The overall style of the video seems to be informative or educational, with the man providing a fact or explanation related to the Coca-Cola can."
+    ],
+    "video_ids": [
+        "SxX58LyX0xg_1_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a blue shirt, bald, wearing glasses; Can of Coca-Cola? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a bald man wearing glasses and a blue shirt, which matches the description. Additionally, a can of Coca-Cola is clearly visible on the left side of the screen. These core elements are present and do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man man wearing glasses and a blue shirt, which matches the description. Additionally, there can of Coca-Cola is prominently visible in the right side of the frame, The elements elements of present and align not contradict the given.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2f63003d9e6b8599125d5bda89b0d6ec10bda085
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0ff43273ddf74ecb9f7c2e82f13267f53b584f4f5282f45f4b894a5fe21aed96
+size 154865
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..042d52b27e2780c62f5fcda5f2c463f470ff64aa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen outdoors, dressed in a casual style. He is wearing a black baseball cap, sunglasses, and a gray hoodie. The man is holding a small object in his hands, which he appears to be examining closely. The setting is a lush green garden with well-maintained hedges and trees. The man is standing near a wooden fence, and the overall atmosphere of the video is relaxed and peaceful. The video captures the man's curiosity and interest in the small object he is holding, suggesting a sense of discovery or exploration. The outdoor setting and the man's attire suggest a casual, leisurely day spent in nature."
+    ],
+    "video_ids": [
+        "9tDQx_Yp8WE_47_0to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a black baseball cap, sunglasses, a gray hoodie, and a small object being held and examined closely.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black baseball cap, sunglasses, and a gray hoodie, and he is holding and examining a small object with his hands. All elements described in the 'Object(s)' condition are clearly present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black baseball cap, sunglasses, and a gray hoodie. which he is holding and examining a small object closely close hands. The the of in the 'Object(s)' condition are present present in match with the video content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e498e1d3a0a10c62109d69377a3c7160416b289b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b480e8b3d80601a0fdca43e0245e039020db9b332c69d5075be92822192ccafd
+size 162121
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..d30949f455402edd2f1ad2d5735800e94d2d838f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a delicious-looking burger being prepared. The burger starts off with a bun, then a layer of sauce is added, followed by a patty, and finally a slice of tomato on top. The burger is placed on a white plate, and there's a spoon in the background. The style of the video is simple and straightforward, focusing on the food and the process of its preparation. The colors are vibrant, and the textures are well-defined, making the burger look very appetizing. The video is likely meant to showcase the burger's ingredients and the care that goes into its preparation."
+    ],
+    "video_ids": [
+        "slp0laqCC0U_82_0to140"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Burger (bun, patty, sauce, tomato slice), spoon? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a burger being assembled with a bun, a patty, sauce, and a tomato slice, which matches the 'Object(s)' condition. A spoon is also visible in the background, though not actively used. No conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close with prepared with a bun, a patty, a, and a tomato slice. which are the 'Object(s)' condition. Additionally spoon is also visible in the background, fulfilling it directly used in The additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b8a278eaa9eee23110c448ffe4118adcc82b5abc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1269-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5ec897ad888d7ba884c929c21a551166d5eb60e7531611c065c540d6d28c19ff
+size 136556
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..56e7ddba8de16d861e6c74c0a7c2945e01e1eb1b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a classic car, showcasing its vintage design and features. The car's dashboard is adorned with a variety of gauges and dials, each serving a specific purpose in monitoring the vehicle's performance. The steering wheel, a prominent feature in the center of the dashboard, is made of wood and metal, adding a touch of elegance and sophistication to the overall design. The car's interior is bathed in sunlight, highlighting the details of the dashboard and steering wheel. The video is a testament to the craftsmanship and attention to detail that went into designing this classic car."
+    ],
+    "video_ids": [
+        "x7skkkY-xq4_50_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dashboard gauges and dials, steering wheel? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the dashboard gauges and dials, along with the steering wheel, which are the primary objects mentioned in the condition. The camera focuses on these elements, and they are prominently displayed throughout the video, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a dashboard gauges and dials, as with the steering wheel, which are the core objects mentioned in the description. The image angle on these elements, providing there are prominently displayed in the frames. fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6d40de28b936f225d2b56eb59ac369d497f1011c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c902d68a74617e4c7d37b2c9940fc07d2e75c817790ad44de3d7526b588e4a72
+size 136372
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..216a756e7a8ec3c45ca1244277edb30f91fb3844
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man in a blue medical uniform is seen in a kitchen setting. He is holding a small object in his hand, possibly a pill or a small device. The kitchen is equipped with blue cabinets and a white countertop. On the countertop, there is a laptop displaying a medical-related website. The man appears to be focused on the object in his hand, possibly examining it or preparing to use it. The overall style of the video suggests a narrative related to healthcare or medical research."
+    ],
+    "video_ids": [
+        "DGq-K8uktOc_115_40to204"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue medical uniform, a small object (possibly a pill or a device) held in his hand, and a laptop displaying a medical-related website.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue medical uniform, holding a small object in his hand, and a laptop displaying what appears to be a medical-related website. All elements described in the condition are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a blue medical uniform, holding a small object in his hand, and there laptop displaying a appears to be a medical-related website. The these match in the condition are present and match with the video content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7817f8321b5e84509e73adcfad38ab27e73969eb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f03872a115a092f0af64796dac68ebb8dbfe1b3448de23da217842f160ed8b9
+size 155507
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..9d2e34e070422cc35197c7d1a9bacfaf2af1bb56
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the serene beauty of a tropical beach. The first frame shows a clear blue sky above the turquoise ocean, with a small island visible in the distance. The second frame shows the ocean gently lapping against the white sandy beach, with a few palm trees scattered along the shore. The third frame shows the ocean's surface shimmering under the sunlight, with a few small waves breaking near the shore. The overall style of the video is calm and peaceful, with a focus on the natural beauty of the beach and ocean."
+    ],
+    "video_ids": [
+        "-KufBNhw2oM_68_0to126"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Ocean, Sky, Island, Palm Trees? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing an ocean with turquoise waters, a blue sky, an island with lush greenery, and palm trees along the shoreline. All specified elements are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting the ocean with varying waters, a sky sky, an island in a greenery, and palm trees. the beach. The these elements are present and contribute depicted,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5bdf9c73ac6c54323a42b5b9b4f1bfe705dfacd5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4ca65f4ad538b75dc663920af9b84d7f0de35324d95c37057b2ef87d4041ab0b
+size 136524
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..e79af6248c4b97aee9f1300ddcb9b98e15be20f2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a motorcycle engine, focusing on the chain and sprocket assembly. The style of the video is a time-lapse or stop-motion sequence, capturing the movement of the chain and sprocket in three different positions. The engine is predominantly silver, with black and orange components. The chain is in motion, moving from one sprocket to another. The sprockets are attached to the engine with bolts, and the chain is secured with a black tensioner. The video provides a detailed look at the mechanical components of a motorcycle, highlighting the precision and engineering involved in its construction."
+    ],
+    "video_ids": [
+        "Yfq_vGV27gY_7_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chain, sprocket assembly, engine (silver), black and orange components, bolts, black tensioner? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a motorcycle's chain and sprocket assembly, a silver engine block, black and orange protective components, visible bolts securing parts, and a black tensioner. All elements described in the condition are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a chain's chain and sprocket assembly, with silver engine,, black and orange components components, and bolts, the, and a black tensioner. The these described in the condition are present and accurately depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0b4e999f8dec5bf52815afe1bb2e60d4e89ea139
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8407da7785ee8202d64b54fbff9d42b840a4581e23ffc990e79614325879ddb3
+size 168661
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..20274afdd2d24caa138d49f56cda5b715d48e9d5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene sailing experience on a clear day. The main focus is a white sailboat with a blue canopy, gliding across the deep blue water. The boat is equipped with a tall mast and multiple sails, which are not visible in the image. The boat is moving towards the right side of the frame, suggesting a journey across the vast expanse of the ocean.  In the background, a majestic mountain range stretches across the horizon, adding a sense of depth and scale to the scene. The mountains are covered in lush greenery, contrasting beautifully with the blue of the sky and the water.  The sky above is a clear blue, with only a few wispy clouds scattered across it. The sun is shining brightly, casting a warm glow on the entire scene and highlighting the details of the boat and the mountains.  Overall, the video presents a tranquil and picturesque sailing experience, with the sailboat as the main subject, set against the backdrop of a stunning natural landscape. The absence of any other objects or people in the frame further emphasizes the solitude and peace of the sailing journey."
+    ],
+    "video_ids": [
+        "FdFK0KcOyDs_58_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white sailboat with a blue canopy, a tall mast, and multiple sails (not visible). A majestic mountain range in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white sailboat with a blue canopy and a tall mast, which matches the description. The background features a majestic mountain range, also matching the description. Although the sails are not visible, this is acceptable as per the condition. The presence of other boats and water is not contradictory to the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a white sailboat with a blue canopy and a tall mast, which align the description. The background features a majestic mountain range, also align the description. The the sails are not visible, the does acceptable as the the condition, The presence of the elements in the reflections not a to the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c69aea095c96585d0f70e077946a1f2d232e342b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:105b69cb6cf7fcd2e4823c79a0d46923e8ce2898fce791fae18850ba9866109e
+size 275653
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..5e6b6deeda6552f350fbc010ce015f0fdcd94df0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a thrilling whitewater rafting adventure on a river. The first frame shows a group of adventurers in their raft, navigating the fast-flowing waters. The second frame shows the raft making a sharp turn around a large rock, with the rafters skillfully maneuvering their craft. The third frame shows the raft emerging from the rapid waters, with the rafters celebrating their successful run. The video is shot from a high angle, providing a bird's eye view of the action. The river is surrounded by lush greenery and towering mountains, adding to the sense of adventure and excitement. The rafters are dressed in safety gear, including helmets and life jackets, emphasizing the importance of safety in such extreme sports. The video is a dynamic and exciting portrayal of outdoor adventure and teamwork."
+    ],
+    "video_ids": [
+        "3Q8mfLzOjYQ_12_399to580"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A raft, a group of adventurers (rafters) in the raft, large rocks in the river.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple kayakers (not a single raft) navigating a river with large rocks, which matches the description of 'a group of adventurers (rafters) in the raft, large rocks in the river.' Although the term 'raft' is not precise (they are in kayaks), the core elements of adventurers on watercraft and large rocks in the river are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a framesaks ina a raft raft as navigating through river with large rocks. which align the description of aA raft of adventurers (rafters) in the raft, large rocks in the river.' The the video 'raft' is used used foras are kay kayaks), the presence elements of adventurers in a with and the rocks in the river are present.}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..eff16028b2d268ef57a829bb3c16750efc4a4008
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:639bf1a521b09ee4cac04b7f15061bcf3713f379fc914d8f1fb72f03eb9eab24
+size 176915
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..ea01029e629b27b662fa70e138271928011fbf49
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a dark shirt sitting in front of a whiteboard with various notes and diagrams. He appears to be in a classroom or office setting. The man is looking to his left, possibly at a monitor or another person. The whiteboard has writing on it, including the words \"Te Bun\", \"Nanikai\", \"Tecorore\", \"Coral Survey\", and \"Ammunition\". There are also several bottles visible in the background. The style of the video is a straightforward, candid shot, likely taken during a meeting or lecture."
+    ],
+    "video_ids": [
+        "TZ0j6kr4ZJ0_307_32to172"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a whiteboard, several bottles, and notes/diagrams on the whiteboard.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man speaking, a whiteboard with handwritten notes, and several bottles on a surface behind him. The whiteboard contains visible text and diagrams related to marine ecosystems and coral surveys, matching the description. No conflicting elements are present that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man sitting in a whiteboard with various notes and and several bottles on the table in him. The presenceboard contains diagrams notes and diagrams, to the biology, biodiversity reefs, which the description. The additional elements are present.\"\n would the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5249f6c60f7341758fed730facffc36e2b8f0e3e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d444e67d0728dda228bc62ac45047f578a5f4b9af613894f64ac195d11c58559
+size 181998
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..143d4fecd302ffeaf9b0308d6620870578db5229
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man with a beard and a baseball cap is seen working on a project. He is wearing a plaid shirt and is holding a power drill in his hand. The man is standing in front of a wooden wall, which serves as the backdrop for the scene. On the table in front of him, there are various tools and materials, including a bottle of glue and a power drill. The man appears to be in the process of assembling or repairing something, as he is actively using the power drill. The overall style of the video is casual and informal, with a focus on the man's hands-on work."
+    ],
+    "video_ids": [
+        "EhhInGkRrDY_6_60to234"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and baseball cap, a power drill, a plaid shirt, various tools and materials (including a bottle of glue) on a table.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and baseball cap wearing a plaid shirt, sitting at a table with a power drill, a bottle of glue, and other tools/materials. These elements match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man with a beard and a cap, a plaid shirt. working at a table with a power drill and a bottle of glue, and other tools ands. The elements match the description provided, the questionObject(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d8d881387498f8d40088bdb52ec4a1e0f925bcb9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-127-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9fd106fe15a614ac8c17e6a95b6f8aea9cb2b15365618aaf4c503c04be8d7b49
+size 118336
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..1467cf8cbcf9ec089e61ef9d5a7df43698319ada
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a chef in a professional kitchen, preparing a dish. The chef is wearing a white chef's coat and is holding a large silver bowl. The kitchen is equipped with stainless steel appliances and counters, and there are various ingredients and tools visible, including a bowl of apples, a cutting board, and a knife. The chef appears to be in the process of mixing ingredients in the bowl. The style of the video is a close-up, real-time shot, focusing on the chef's actions and the details of the kitchen and ingredients."
+    ],
+    "video_ids": [
+        "7Aw9JF7Mj6M_7_101to238"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chef, large silver bowl, bowl of apples, cutting board, knife? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a chef in a professional kitchen setting, holding and mixing a large silver bowl. A bowl of apples is visible on the counter to the left, and a cutting board with a knife is also present on the counter. All specified objects are present and correctly identified in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a chef in a professional kitchen environment. using a using ingredients large silver bowl. There bowl of apples and visible on the counter, the left, and a cutting board with a knife is present present. the counter. The the objects are present and match described in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..657d9ad31c96d882134c976182c0ae63a8d0842a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:abde6952717fbdfe3438087535a9b341ff3221106d1c1271c7ed2db6cada72a3
+size 175890
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..b219fa9992983519add25acf1603db97f185189e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a black and tan Doberman Pinscher stands on a wooden porch, looking directly at the camera. The dog is wearing a black collar with a tag. In the background, there is a white door and a pair of black boots. The porch has a brown rug on it. A few moments later, an orange tabby cat appears on the porch, lying down and looking at the camera. The cat is near the rug. The video captures the interaction between the two animals in a calm and peaceful setting."
+    ],
+    "video_ids": [
+        "ZBlh77QCAIs_16_0to177"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black and tan Doberman Pinscher and an orange tabby cat? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a black and tan Doberman Pinscher sitting on a porch, and an orange tabby cat lying nearby. Both animals are accurately represented as described in the condition. The background elements (like boots and a door) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a black and tan Doberman Pinscher and on a wooden, and shortly orange tabby cat appears down. The animals are present depicted, described, the condition, The presence and,por the and a door) do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d04bf3f5305ed08de68b8588b3724b2e81e8f0e5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f421d269e059ec14c747597bc15eccd4f059eb2bea675e5c9144a5e36142ae3b
+size 150379
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..030692e6c1706e8bba9d06fe736ab2057e0e8209
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in front of a silver car, likely a car salesman or a car enthusiast. He is wearing a black polo shirt with a red logo on the left chest, suggesting he might be associated with a car dealership or a car-related event. The man is wearing glasses and has short, light-colored hair. The car is a sleek, modern design, possibly a luxury or sports car, and it is parked in front of a building with large windows. The setting appears to be an outdoor car show or a car dealership, as indicated by the presence of the car and the man's attire. The style of the video is straightforward and informative, likely aimed at showcasing the car or the dealership."
+    ],
+    "video_ids": [
+        "Iu96bgqURZc_2_19to150"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a silver car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man standing in front of a silver car, which fulfills the 'Object(s)' condition. The background elements, such as the building, do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man standing in front of a silver car. which fulfills the 'Object(s)' condition. The man includes, such as the building and do not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..804983b41fea91bc3d43286140a549dbb3820b0e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6e925d1bc1ab4dce6667af15551d4d3f59f91eee213265e0d53f056a3b45aab6
+size 168264
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..505d9f192e39ec7ffd9891274af28ce786bd1594
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of two birds in flight against a blurred natural background. The birds, likely doves or pigeons, are shown in mid-flight, their wings spread wide as they soar through the air. The focus is primarily on one bird, which dominates the frame, while the second bird appears slightly out of focus in the background. Both birds exhibit a mix of white and gray feathers, with darker markings visible on their wings and tails. The lighting suggests it might be late afternoon or early evening, casting a soft glow on the birds' feathers. The camera remains steady throughout the sequence, maintaining a consistent angle that emphasizes the birds' graceful movement and the fluidity of their flight. The overall scene conveys a sense of freedom and tranquility, highlighting the beauty of nature in motion."
+    ],
+    "video_ids": [
+        "d08953fb0bb912c66225a2d1ba39eb1abf822096b0ebddbd296425555d5488bb"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two birds, likely doves or pigeons.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two birds in flight, with visible wing movements and body shapes consistent with doves or pigeons. Their plumage appears to be primarily gray and white, which matches typical dove/pigeon coloring. The motion and context strongly support the description without contradicting it.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two birds in flight, which their wings patterns and a shapes that with doves or pigeons. The plumage and to be a white and white, which is the dove origeon color. The presence and posture suggest suggest the identification of anying it.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3b1515f0b7c57a8d765c1ea48df90f6a8d3ddd03
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b50ffbdf3f995378bd97672375076b324e57e4296f4eb2795482f766a7583d8
+size 171184
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..969e145c31fca207fa468b3023bf5f0b79bc834d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures an exhilarating moment of a surfer riding a powerful wave. The camera angle is from above, providing a bird's-eye view of the action. The wave is a vibrant blue, with its crest curling over, creating a dramatic arc. As the surfer maneuvers skillfully on their board, they maintain balance amidst the wave's forceful movement. The water splashes around them, adding to the dynamic energy of the scene. The sunlight reflects off the surface of the water, creating a shimmering effect that enhances the visual appeal. The surfer's body is partially submerged, showcasing their agility and control as they navigate the wave. The overall atmosphere is one of adventure and the raw power of nature."
+    ],
+    "video_ids": [
+        "9bce0570f982976d24ea39b91e4c9b51c71418f870cd1b64600149d85f63c527"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Surfer on a board, wave, splashing water? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a surfer on a board riding a wave, with visible splashing water. The surfer is partially obscured by the spray but identifiable, and the wave and splashing water are clearly depicted, fulfilling the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a surfer on a board riding a wave, with spl splashing water around The surfer is depicted submerged by the wave, is by and the wave and splashing water are clearly depicted, fulfilling the ' description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f82ffdccf473a1da5b199b4f824f8f1fafc43473
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0fd90fe61d20d1b1352156046b5850ca06ffb905070af5926831b5a9dd78e24d
+size 281631
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..908f66b722eb4f38d904ded6f18d5950a839d3b0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car, focusing on the center console and the gear shift. The car has a luxurious design with tan leather seats and a black center console. The gear shift is made of black leather with silver accents. The center console has a silver panel with multiple buttons and a red button on the left side. The car's interior is well-lit, and the sunlight is shining through the windows, creating a warm and inviting atmosphere. The video is a close-up shot, focusing on the details of the car's interior, and it is likely taken from the driver's perspective. The style of the video is realistic and detailed, capturing the textures and colors of the car's interior with precision."
+    ],
+    "video_ids": [
+        "NtDj3W6GAn4_27_0to174"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Center console, gear shift, leather seats, silver panel with buttons, red button? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows the center console with a gear shift, leather seats, a silver panel with buttons, and a red button. All these elements are clearly visible and match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a center console, a gear shift, leather seats, and silver panel with buttons, and a red button. The elements elements are present visible and match the description provided.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ec159e44317f922dc73f3c3d4a6c5f39689c175e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37a74290b5ace870475dfb9161e31ed764712c89f0d1fcaf25d9a5323254d4ee
+size 99799
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..0af6fb3bd7d26546e852552a3ab214320595fe16
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a modern, luxurious living space with a panoramic view of a city skyline. The living room is furnished with contemporary pieces, including a white sofa, a gray armchair, and a sleek coffee table. The room is illuminated by natural light streaming in through the floor-to-ceiling windows, which offer a stunning view of the cityscape. The glass walls and railing of the balcony provide an unobstructed view of the city, while the curved architecture of the building adds a unique aesthetic to the space. The overall style of the video is sleek and sophisticated, highlighting the elegance of the interior design and the breathtaking view of the city."
+    ],
+    "video_ids": [
+        "ZXngwK_f2yQ_42_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: White sofa, gray armchair, sleek coffee table, floor-to-ceiling windows, glass walls, balcony railing, curved building architecture.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a modern, luxurious interior with a white sofa, gray armchair, and sleek coffee table visible through the floor-to-ceiling windows. The building features glass walls and a balcony railing, and its architecture is distinctly curved. All elements described in the condition are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a modern living high living with a white sofa, a armchair, and a coffee table.. the large-to-ceiling windows. The windows's glass walls and a balcony railing, and the architecture includes curved curved, The elements in in the condition are present and match depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..830e8a9f3aab752ef05953ee44fcf66dd18f8e6a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:db6fd055f0400590997dc2eebdae60c76e0a57a28a70da71019bb00703ca8a10
+size 121461
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..fc74aaa060b6c641f006d1716f86e76af846cf82
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features an elderly woman sitting in a chair, engaged in a conversation. She is wearing a blue jacket and has a microphone attached to her collar, suggesting that she is participating in a broadcast or interview. The woman is gesturing with her hands as she speaks, indicating an animated discussion. In front of her is a glass of water, which she appears to be sipping from. The setting is a simple, uncluttered room with a white wall in the background. The overall style of the video is straightforward and focused on the woman and her conversation, with no additional elements or distractions."
+    ],
+    "video_ids": [
+        "OUKtla7ItAE_12_78to265"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An elderly woman, a glass of water.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an elderly woman seated at a table with a glass of water in front of her. Her appearance and the presence of the glass of water align with the specified 'Object(s)' condition. No elements contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features an elderly woman sitting at a table with a glass of water in front of her. The attire and the setting of the glass of water align with the description 'Object(s)' condition. The additional contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d2133ec8b3af4edef64ad1c156562a48f003e40a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1270-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e14fdec036bc7b94764d53c8cc1ebbd94c38bb96b41c553030ec29b8bb4118e
+size 210785
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..7223a98c85c427dcc86745c394470b1eb9da5006
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene indoor scene featuring a black cat exploring a sunlit patch on a carpeted floor. The room is minimally furnished, with a wooden bed frame holding a blue and white checkered blanket, a small nightstand with a lamp, and a trash can near the wall. The sun casts a rectangular beam of light onto the floor, creating a stark contrast between the illuminated area and the surrounding darkness. The cat, initially positioned at the edge of the light, moves gracefully across the beam, its silhouette sharply defined against the bright background. As it continues to explore, the cat's movements are fluid and curious, occasionally pausing to investigate the light. The camera remains stationary throughout, focusing on the cat's interaction with the light, highlighting the interplay of shadow and light in the quiet, domestic setting."
+    ],
+    "video_ids": [
+        "babafcaf2fa308dd55b451b2e955be3b252bdfcafccfc62d50a48965432d2a7e"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Black cat, wooden bed frame, blue and white checkered blanket, small nightstand with a lamp, trash can? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black cat moving around a room with a wooden bed frame, a blue and white checkered blanket on the bed, a small nightstand (though no lamp is visible on it), and a trash can near the nightstand. All these elements are present and match the description, even if the lamp is not visible on the nightstand.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black cat walking across in room. a wooden bed frame, a blue and white checkered blanket on the bed, and small nightstand withwhich the lamp is visible), it), and a trash can. the bedstand. The the elements match consistent and match the description provided fulfilling though the lamp on not visible, the nightstand.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d5483fbd1559eaf85aa7adc47852219cedb36e5d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d724b03a8f6b46e70a4338705e7209cb9d83686c8cffa45d638d8275d38aca46
+size 133211
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..8f8ec0f7fe3845c0ad0e84bb3ee5ea379c3a592e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a man and a woman in a studio setting. The man is wearing a suit and tie, while the woman is dressed in a blue dress. They are seated at a white desk with a laptop and a few papers on it. The background is purple with a large window that offers a view of the city. The man and woman are engaged in a conversation, with the woman gesturing as she speaks. The overall style of the video is professional and polished, typical of a news broadcast."
+    ],
+    "video_ids": [
+        "ch6tTpX-3HM_13_0to164"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, a woman in a blue dress, a white desk, a laptop, and some papers.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit and tie, a woman in a blue dress, a white desk, a laptop, and some papers, which matches the core description. The setting is a news studio, and all specified objects are clearly visible and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man in a suit and tie and a woman in a blue dress, a white desk, a laptop, and some papers. which align the description description. The presence appears a studio studio with and the the elements are present visible and in with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..81995a599c0be182aca05e040e4fc50d5eafc584
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b110ef0fde72f572c8f623168c9007ad96a55c1539d43b095d05408ffe9ee13
+size 181635
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..501477c06458faa39d1ebb874b41f67c82d4cc70
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of triumph for a football player. The player, wearing a vibrant green and yellow uniform with the number 12, is in the midst of a celebration. He is on the field, surrounded by a crowd of spectators, and his arms are raised high in the air, a universal gesture of victory. The player's helmet, matching the color of his uniform, is clearly visible. The crowd in the background, though blurred, adds to the atmosphere of the scene. The video is a dynamic snapshot of a moment of triumph in the world of football."
+    ],
+    "video_ids": [
+        "EY4sihN6UJc_52_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A football player in a green and yellow uniform with the number 12, a helmet matching the uniform color.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a green and yellow uniform with the number 12, and a helmet that matches the uniform's color scheme. The player is clearly identifiable as a Green Bay Packers player, which aligns with the description. The background and additional elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a green and yellow uniform with the number 12 prominently and a helmet that matches the uniform color color.. The player is also in as a football Bay Packers player, which aligns with the description. The presence and additional elements, not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..94fd55827b5ddc44c2f15f626467d8a640154695
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0cf9cacd562073c859462824d28f5def6c01ab73b874adcba7735727ce1062c8
+size 240768
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..5bd4436517533fde89dc0e0cc9d478574feb4ba9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen walking through a large, industrial warehouse. He is dressed in a casual gray t-shirt and is carrying a jacket over his shoulder. The warehouse is filled with tall metal beams and large windows that let in natural light. The man appears to be in motion, possibly walking towards the camera. The overall style of the video is realistic and it captures the man's movement and the industrial setting in a natural and unobtrusive way."
+    ],
+    "video_ids": [
+        "XT2IS2dn8gM_2_46to182"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a casual gray t-shirt and a jacket over his shoulder.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The man in the video is wearing a casual gray t-shirt and has a jacket over his shoulder, which matches the description. The jacket is black leather, and he removes it to reveal the gray t-shirt underneath. The setting and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video in the video is wearing a casual gray t-shirt and has a jacket over his shoulder, which matches the description provided The setting is partially and, and the is it as reveal the t t-shirt.. The setting appears the elements in not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e152ed12446a5e92bdef7ebae83ff92e9dab6822
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c300688b5d29c49d6696e5b17c8eb5352dd5a9c8c5ff7463870a08edc7788b84
+size 237082
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..497b31c381868251cbb487cb1ca80d9a2899e00d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a field blooming with purple flowers. The flowers, in various stages of bloom, are scattered throughout the field, creating a beautiful contrast against the lush green grass. The tall grass sways gently in the breeze, adding a sense of movement to the otherwise still scene. The video is shot from a low angle, making the flowers appear larger and more prominent. The overall style of the video is naturalistic, with a focus on the beauty of nature. The colors are vibrant and the lighting is soft, creating a peaceful and calming atmosphere."
+    ],
+    "video_ids": [
+        "R-gAaA6qGuw_11_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Purple flowers, green grass? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features purple flowers and green grass, which aligns with the 'Object(s)' condition. The purple flowers are clearly visible in the foreground and background, and the green grass forms the majority of the field, fulfilling the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video predominantly features purple flowers, green grass, which aligns with the 'Object(s)' condition. The flowers flowers are the visible and the foreground and background, and the green grass is the backdrop of the background, fulfilling the description description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..96f73044e8c96233e5e83d80e91fcaef7aae3f84
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:515151fa39c44452ccded2daa05749b688bc1f82743f571d4c7bf5047f76374e
+size 255069
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..d3ddfab84a1d545d83d51b131774909fc5965d2b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a chef in a professional kitchen, preparing a dish. The chef is wearing a white chef's coat and glasses, and is focused on the task at hand. The kitchen is well-equipped with stainless steel appliances, including a stove and an oven. There are also several pots and pans hanging from a rack above the stove. The chef is using a wooden spoon to mix ingredients in a bowl. The overall style of the video is realistic and informative, likely aimed at teaching viewers how to prepare a specific dish. The chef's actions are the main focus of the video, with the kitchen serving as a backdrop."
+    ],
+    "video_ids": [
+        "eOV9CmpA69o_30_0to120"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A chef in a white chef's coat and glasses, a wooden spoon, a bowl, pots and pans hanging from a rack.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chef wearing a white chef's coat and glasses, using a wooden spoon to mix ingredients in a bowl. Pots and pans are clearly visible hanging from a rack in the background. All core elements described are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chef in a white chef's coat and glasses, stirring a wooden spoon to stir ingredients in a bowl. Theots and pans are hanging visible hanging from a rack in the background, The elements elements of in present and accurately depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..059e9c9fdfccbd5705de5beb31b499404da81676
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7425a5e38639dd6f04b7752ae7910e30d49738e95e99599f1690c8272cfd173b
+size 216879
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..15db86f3f86568bef44d1c4c930122ffa570c96f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit sitting at a table with a laptop and a plate of food. He appears to be in a formal setting, possibly a conference room or a meeting area, as indicated by the presence of a flag and a chair in the background. The man is engaged in an activity that involves the use of a red object, which he holds in his hands. The laptop is open and appears to be in use, suggesting that the man might be working or preparing for a presentation. The overall style of the video is professional and formal, with a focus on the man and his actions."
+    ],
+    "video_ids": [
+        "6sjkEVUl1ao_18_57to182"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a suit, laptop, plate of food, red object? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit sitting at a table with a laptop and a plate containing a red object, which matches the core description. Additional elements like flags and a grenade do not contradict the presence of these core objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit sitting at a desk with a laptop and a plate of various variety object, which appears the description description. The elements like the in a red are not contradict the core of the core objects.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..599000f03e55585dd3c432190d1c483d024c8cb9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3309edb44502af856f1c05a3bc9d0a9245230ddcc73e1b4e7c8a980718d30893
+size 133532
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..a5a591e63d488c5ab786ac42f0948e367f56ab43
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a 3D animated character, a man with a beard and mustache, who appears to be a doctor or a scientist. He is wearing a white lab coat and a blue shirt. The man is standing in a room with a white wall and a clock on the wall. He is looking to the right with his mouth open, as if he is speaking or reacting to something. In the background, there is a green square with white silhouettes of people, suggesting a group or a team. The overall style of the video is 3D animation with a realistic and detailed rendering. The character's expression and the green square with silhouettes of people suggest a narrative or a concept related to communication, teamwork, or a presentation."
+    ],
+    "video_ids": [
+        "NVc6EapsGtw_2_0to154"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and mustache wearing a white lab coat and a blue shirt, a green square with white silhouettes of people.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man with a beard (and no mustache, but the description may be imprecise) wearing a white lab coat over a blue shirt, which matches the core description. Additionally, a green square with white silhouettes of people appears in the video, fulfilling that part of the condition. The presence of other elements (like changing numbers and a skull icon) does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man with a beard andmust must mustache) but this description does have referringprecise) wearing a white lab coat and a blue shirt. which matches the description description. Additionally, there green square with white silhouettes of people is in the background, which the part of the description. The presence of a elements inlike the facial on a clock)) does not contradict the main description and}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..05106e961d0a695aa85880ca9b9768e716d422d8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1271-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a143dda5edd55d9be6fba4dcc7111dba53e19dd5a5cdc5e5fb5402bc0b6d7e43
+size 179411
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..f9f17b60a9f53cc27193acc2118256d0b1838fd3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the process of pouring honey into a bowl. The honey is golden and viscous, dripping from a spoon held over the bowl. The bowl is white and round, and it sits on a white surface. In the background, there are two stacks of brown bottles, possibly containing more honey or other ingredients. The style of the video is simple and straightforward, focusing on the action of pouring the honey without any additional context or embellishments. The lighting is bright and even, highlighting the golden color of the honey and the white of the bowl and surface. The video does not contain any text or additional elements, keeping the focus solely on the action of pouring the honey."
+    ],
+    "video_ids": [
+        "fYZYMESqjf0_38_0to151"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Honey, spoon, bowl, brown bottles? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows honey dripping from a spoon into a bowl, with brown bottles visible in the background. The scene transitions to a display of honey jars, which still aligns with the presence of honey and related containers. All core objects mentioned are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful the being from a spoon into a bowl, which a bottles in in the background. The objects matches from a close of honey dripping, which align aligns with the presence of honey and brown objects. The elements elements ( in present and correctly depicted.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0cc2f4915c0b6227adf8c3ad7b7aeade407dc98a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b91f266f7ec1d0a87925f53bf21355c07fd6cfb8474c5a24b486702969e10cd
+size 95873
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..dc5372cdc17ca464610d64d2e7e98b44724be518
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a stylized, animated depiction of Captain America, a popular comic book character. The scene is set in a room with a green door and a shelf filled with various bottles. Captain America is shown in three different poses, each more dynamic than the last. In the first frame, he is standing with his hands on his hips, looking confident and ready for action. In the second frame, he is holding a shield, preparing to defend himself. In the third frame, he is in a fighting stance, ready to take on any challenge. The animation is smooth and fluid, capturing the essence of the character's strength and determination. The background is simple, allowing the focus to remain on Captain America. The overall style of the video is reminiscent of a comic book, with bold lines and vibrant colors."
+    ],
+    "video_ids": [
+        "I7fFt7qoCDQ_20_21to193"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Captain America? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features Captain America in a dynamic, comic-style illustration, which matches the 'Object(s)' condition. The surrounding elements (wall, shelves, etc.) do not contradict the presence of Captain America and are acceptable as background context.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a America, a dynamic pose animated book animation. performing align the descriptionObject(s)' condition. The character elements,b of bottles, and.) do not contradict the core of Captain America and are consistent as additional details.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..321f108aa839ac42d574b493ac8c263c275724fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:35d13b9c850b80c1b63dc8b730e4c76139d4c4b88a349ee86b52ecf0ceba9bdb
+size 321681
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..7e21d0de278986fe1816b845643b842007576240
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of celebration between two soccer players on a field. The player in the foreground, with a broad smile on his face, is embracing the player in the background, who is bald and wearing a jersey with the name \"ZIDANE\" and the number \"5\" on it. The players are wearing white jerseys, and the field appears to be wet, suggesting that it has been raining. The background is blurred, but it seems to be a stadium filled with spectators. The overall style of the video is dynamic and captures the excitement of the moment."
+    ],
+    "video_ids": [
+        "_pnkN6PO22g_11_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two soccer players? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two soccer players in white jerseys celebrating on a field. One player has long hair and is smiling broadly, while the other, with a bald head, has his back turned and is raising his arm. Both are wearing jerseys with visible names and numbers, consistent with soccer players. The scene is set in a rainy environment, which is typical for outdoor sports, and there are no elements contradicting the description of two soccer players.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two soccer players in white uniforms, on a field. The player is the hair and the wearing,, while the other is bald a bald head, is his arm to to is embracing his arm in The are wearing soccer with the numbers and numbers, which with the players. The background is set in a stadium stadium, which is a for soccer sports events and the are spectators elements thating the description of two soccer players.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4d3d08ffe8be53b1b368611d3df76535c3ab0d99
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0347aea58cec68163284309fbc4f935dd820d6b089b426592972944bc59162b6
+size 209865
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..d436a8c110040be0be79c30e41f27c01157cd607
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a large metal bowl filled with small, round, yellow objects, which appear to be food items. The bowl is placed on a metal surface, and there are two utensils resting on the edge of the bowl: a wooden spoon and a metal strainer. The style of the video is a simple, straightforward food preparation or cooking video, focusing on the ingredients and the tools used. The lighting is bright, and the focus is sharp, highlighting the texture and color of the food items. The background is minimalistic, with no additional objects or distractions, allowing the viewer to concentrate on the contents of the bowl."
+    ],
+    "video_ids": [
+        "3ubVIdMI6pQ_47_77to217"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large metal bowl, small, round, yellow food items, a wooden spoon, and a metal strainer.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large metal bowl filled with small, round, yellow food items, which appear to be fried or cooked. A wooden spoon is visible, inserted into the bowl, and a metal strainer (or skimmer) is also present. These elements match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large metal bowl containing with small, round, yellow food items, which appear to be some or cooked. There wooden spoon is placed inside resting into the bowl, and a metal strainer islikely similarimmer) is placed present on The elements match the description provided, the questionObject(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ed2e302403f4761da3062a1ec579672d6883abf8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e3b66b24c83ddae3b0c38e137bf0fe36904541b2029f49896baae50524c6c5d4
+size 81582
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..37ef179264f33898fb6eda299edc9c00f4a742e6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman named Sonia Jones, who is the main subject. She is seen in a grey shirt, sitting in a room with a brick wall and a statue in the background. The room has a warm and inviting atmosphere, with a window that lets in natural light. Sonia appears to be engaged in a conversation or an interview, as she is looking to her left with a thoughtful expression on her face. The video seems to be a casual, yet professional, discussion or interview, possibly related to her work or personal life."
+    ],
+    "video_ids": [
+        "FWZBuc13Zuw_4_17to189"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman named Sonia Jones in a grey shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman with blonde hair wearing a grey shirt, and a lower-third graphic identifies her as 'SONIA JONES'. The description matches the core elements of the video without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a woman wearing long hair wearing a grey shirt, which the name-third graphic appears her as SoniaSIA'.ONES'. The setting matches the core elements of the video, any.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9b112b983f3b29eaeea0756c6ffcb28b80b9a5dd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6ca5fa626b1ba9687c60da9eed2eb026484f59d06712ca95be0a30c1acb095e8
+size 195293
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..b25a515b9e4daf017d1e283be56313b9c0764001
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a man in a suit being interviewed by a group of reporters. The man is wearing glasses and appears to be speaking into a microphone. The reporters are holding up their microphones towards the man, indicating that they are recording his statements. The setting appears to be outdoors, as suggested by the natural light and the presence of a fence in the background. The style of the video is a standard news report, with the focus on the man being interviewed and the reporters capturing his words. The video likely contains audio of the man speaking and possibly the questions posed by the reporters."
+    ],
+    "video_ids": [
+        "OQCSXdVri3U_5_0to143"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit wearing glasses, a group of reporters holding microphones.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit wearing glasses, surrounded by reporters holding microphones. The scene matches the described 'Object(s)' condition, with no conflicting elements. Additional elements such as smartphones and other reporters are present but do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man in a suit wearing glasses, and by a holding microphones. This man is the description elementsObject(s)' condition without with the additional elements present The elements like as the and a reporters are present but do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4220c74f72ad60499b3f207097b632e4e2487930
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17b5affc1ad2a627b9c44b7869c10b97f4a03e331bfdd78a25a706aff2b559bc
+size 160986
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..b0a22080f746933c9a68d013fb9781825a0f6485
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the growth and development of a fruit tree over time. The first frame shows a small, red fruit beginning to form on a branch. In the second frame, the fruit has grown larger and is starting to ripen. By the third frame, the fruit has reached its full size and is fully ripe, ready for picking. The tree's leaves and branches also change over time, providing a sense of the tree's growth and development. The video is a close-up shot, focusing on the fruit and its immediate surroundings, and is shot in natural light, highlighting the colors and textures of the fruit and the tree."
+    ],
+    "video_ids": [
+        "kr5XWfXselw_61_0to123"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Small, red fruit, larger ripening fruit, fully ripe fruit, tree's leaves and branches? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three small, red fruits on a branch, with one appearing slightly larger and more mature than the others, suggesting a ripening process. The surrounding foliage includes tree leaves and branches, which are clearly visible and consistent with the description. There are no conflicting elements that contradict the specified objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a stages red red fruits hanging a tree, with one larger larger larger, another rip than the others, indicating a ripening process. The background environment consists tree leaves, branches, which are consistent visible. match with the description. The are no additional elements in would the core conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a7878435d265f07b5ecbcf828013d1a01ec9c427
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60f8cc8fc0a3b791e822f9c1488676938b19b43d5fc05802d2e0b6a1ea6b063e
+size 220094
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..cca389275d68b61f72b052ab2f75bea412acad25
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red car parked in a garage. The car is shiny and appears to be in good condition. The garage is spacious and well-lit, with a concrete floor that reflects the car's red color. In the background, there is a white airplane with a red star on its tail, suggesting that the garage may be located in a country with a communist history. The airplane is parked on a tarmac, indicating that the garage is likely located near an airport. The overall style of the video is realistic and documentary, capturing the details of the car and its surroundings with clarity and precision."
+    ],
+    "video_ids": [
+        "k0sDY_jK5R4_36_29to214"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red car parked in the garage and a white airplane with a red star on its tail.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red car parked in a garage, and a white airplane with a red star on its tail is visible in the background. These objects are prominently featured and match the description provided. Additional elements like boxes and equipment in the background do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a red car parked in a garage, which there white airplane with a red star on its tail is visible in the background. The elements match consistent featured and match the description provided. The elements, the and a in the garage do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1feb686da9ef3c14358ed5625c8d4f8800501462
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1272-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03df25cb20f0430690f9e16470e2b6c8f1897f26c61277b1d2a8ed98b508453b
+size 128072
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..cd8b10ae2c011381c9b882bb7695c38ff55de340
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is an aerial view of a luxurious estate, showcasing its various amenities and the surrounding landscape. The estate features a large, well-maintained tennis court, a swimming pool with a curved shape, and a spacious lawn. The architecture of the main house is modern and elegant, with multiple levels and a mix of flat and sloped roofs. The property is surrounded by lush greenery, including palm trees and other tropical plants, which contribute to the overall serene and upscale atmosphere. The aerial perspective provides a comprehensive view of the estate's layout and design, highlighting the thoughtful integration of the man-made structures with the natural environment."
+    ],
+    "video_ids": [
+        "5a3v7vCCUdM_4_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Tennis court, swimming pool, lawn, main house? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a tennis court, a swimming pool, a well-maintained lawn, and the main house, all of which are prominently featured in the aerial view. These elements are consistent with the requested 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a tennis court, swimming swimming pool, a lawn-maintained lawn, and a main house. which of which are consistent featured in the scene view of The elements match consistent with the ' objectsObject(s)' and.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e815ebd8ae0b12d26387da6a79a5f25c38782119
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:387bb6676d46f80b8a2555d5f6434da20d0f1ebb5d270807bc1cddf0838aaf01
+size 164196
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..2fa680293c2870337414db30407b8b0a0587fac4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a red SUV in a studio setting. The car is positioned in the center of the frame, with a black background that contrasts with the vehicle's vibrant color. The SUV is captured from a three-quarter angle, highlighting its sleek design and modern features. The car's large wheels and black rims add to its sporty appearance. The studio lighting casts a soft glow on the vehicle, accentuating its curves and contours. The overall style of the video is sleek and modern, with a focus on the car's design and features."
+    ],
+    "video_ids": [
+        "C03ZOF3hwVg_1_0to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red SUV? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red SUV, specifically the NIO ES6 2019 model, which matches the description. The vehicle is clearly visible and is the main focus of the frame, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a red SUV, which a frontIO ES8,2029 model, which matches the description of The vehicle is the visible and is the main focus of the video. fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6a9a9d961f34b7c571a0bd95e31c627478e4d064
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e12afb0673c72aa7ae529f74a012c66a1a336b32ed0a77a69cf175aaaee51a2f
+size 87448
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..7c9c080f0018608f998c98799ecc5e15588aa617
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a small bird perched on a mossy log. The bird, with its gray and orange plumage, stands out against the green moss. The log, covered in moss, provides a natural and textured surface for the bird to perch on. The bird's position on the log suggests it might be resting or observing its surroundings. The background is blurred, drawing focus to the bird and the log. The overall style of the video is a close-up shot, emphasizing the bird and its immediate surroundings. The video does not contain any text or additional objects. The bird's actions are limited to perching and possibly observing its surroundings. The video does not depict any interactions between the bird and other objects or creatures. The focus is solely on the bird and its immediate environment."
+    ],
+    "video_ids": [
+        "1qLf5ZDy0cM_9_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small bird with gray and orange plumage.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small bird with distinct gray and orange plumage, consistent with the description of a robin. The bird's coloration and size match the specified criteria throughout the video frames.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a small bird with gray gray and orange plumage, which with the description. a bird. The bird is coloration and size are the characteristics ', the video.\"\n.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..35291d9985658cd4890b0c95bcb3940fc380a755
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e806bcc97724d9a175cdfb7e558892cca81fa05bc4220b38cd5320ac093cfcb8
+size 173849
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..dc9a2367bd4d9d735191f1e2b9fdbc377a1a94e1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene journey on a boat across a vast body of water. The boat, with its white hull and green canopy, is the main focus of the video. It's equipped with a tall mast, suggesting it's a sailboat. The boat is seen from a high angle, providing a bird's eye view of its journey.  The journey begins with the boat sailing towards a small island. The island, covered in lush greenery, stands out against the vast expanse of the water. The boat continues its journey, moving further away from the island.  As the boat sails on, the sky above becomes overcast, casting a soft light over the scene. The water around the boat is calm, reflecting the tranquility of the journey. The boat, the island, and the overcast sky create a harmonious scene of a peaceful day out on the water."
+    ],
+    "video_ids": [
+        "LVmhSMbGUag_8_0to221"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white hull boat with a green canopy and a tall mast, a small island? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a white-hulled catamaran with a green canopy and a tall mast, consistent with the description. It is also sailing near a small, green, tree-covered island, matching the second part of the description. The overall scene aligns well with the specified objects without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a white hullulled boatamaran with a green canopy and a tall mast, which with the description. Additionally is also situated near a small island green-covered tree-covered island, which the second part of the description. The presence scene aligns with with the given objects.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..329c021c7a98d6a4820e05a41e0dffe19c8ab50e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dafd5a64f42680b589eabae90c6824e707a55fb9d56046ceebc2b16764753616
+size 202831
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..e7039b8efc6520a7b56525f1e1159cee96348599
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene bedroom scene. The room features a neatly made bed with blue bedding and a wooden headboard. A small nightstand sits beside the bed, holding a few items, including what appears to be a lamp and some papers. A trash can is positioned near the wall, and a vent is visible beneath it. The walls are painted a light beige color, and a window with blinds allows natural light to filter into the room.  As the video progresses, a black cat enters the frame from the right side, walking towards the center of the room. The cat's movement is smooth and deliberate, adding a dynamic element to the otherwise static scene. The cat continues its journey across the carpeted floor, eventually exiting the frame on the left side. The overall atmosphere remains calm and quiet throughout the video, with no other significant changes or movements occurring."
+    ],
+    "video_ids": [
+        "58cd0fd5242df95e578e227f5116725c3ed0a48e1f303ebde7e92124503cc791"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black cat, a small nightstand, a lamp, some papers, a trash can, and a vent.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a black cat, a small nightstand, a trash can, and a vent, as described. Although a lamp and papers are not explicitly visible, their absence does not contradict the core description, and the presence of additional elements (like the cat's movements) is acceptable. The scene largely matches the requested objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a black cat, a small nightstand with a lamp can, and a vent. which described. The there lamp and some are not explicitly mentioned, the presence does not contradict the core description, as the presence of the elements likelike the bed and shadow) does acceptable.\"\n The video is matches the given objects.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9d7bfc3f23f65a62009c1784cd6dd68aef440f1e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:713741891e5996acf16ad6921e328da9d309624f52ece76680fea27b97a33c94
+size 234455
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..7ae1520973593534305baf98e427938e58af37c1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a collection of baskets filled with fresh, colorful bell peppers. The baskets are arranged in a row, each containing a variety of bell peppers in different shades of green, red, and yellow. The peppers are ripe and appear to be freshly harvested, with some still retaining their green stems. The baskets are placed on a flat surface, possibly a table or a counter, and the lighting suggests an indoor setting. The style of the video is straightforward and documentary, focusing on the presentation of the bell peppers without any additional context or narrative. The video is likely intended to showcase the fresh produce for sale or for use in cooking."
+    ],
+    "video_ids": [
+        "5GsfU63pZpw_55_0to148"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Baskets filled with fresh, colorful bell peppers in different shades of green, red, and yellow.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple baskets filled with fresh bell peppers in various shades of green, red, and yellow, matching the described condition. The peppers appear vibrant and freshly harvested, and the baskets are clearly visible. There are no conflicting elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows baskets baskets filled with bell, peppers in various colors of green, red, and yellow. which the description '. The peppers are vibrant and are picked, and the baskets are clearly visible, There are no additional elements in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6afe7a16aa876b0ab1efba000ab426843ccdc3e5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6f54b88dee8872a62eff433f812569bf3ec084f3a3409a71bbb93f7fd6eec7f7
+size 107976
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..6d9afbe7ef8ce2074940219394fe1845171f4e14
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a unique architectural structure with a distinctive roof design. The roof is made of wooden shingles, giving it a rustic and natural appearance. The structure has a pointed roof, which is a common feature in traditional European architecture. The roof is also adorned with a small window, which is situated in the center of the roof. The window is framed in red, adding a pop of color to the otherwise neutral tones of the roof. The structure is surrounded by trees, suggesting that it is located in a wooded area. The overall style of the video is a blend of architectural and natural elements, creating a harmonious and visually appealing scene."
+    ],
+    "video_ids": [
+        "avn7brkxmmo_33_0to180"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A structure with a pointed roof and a small red-framed window on top.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a structure with a pointed roof, and a small red-framed window is visible near the top. The roof's pointed shape and the window's red frame match the description, even though the structure has an unusual, organic design and additional elements like shingles and surrounding trees.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a structure with a pointed roof and which there small window-framed window is visible on the top of The roof is pointed shape and the window's red frame are the description provided and though the window appears a unusual design almost texture that is elements like theingles and a foliage are}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..226220bf2784e991a579cd9069fb4affd7f91b6f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c24e7027ab0e5721187f5e33f33d37945a184317d904d2cd169e4979f82cf6f
+size 171136
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..585cf346c263f60d91d47440fbb605a1a483f7cc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up of a bird with a striking appearance. The bird's head is adorned with a vibrant blue crest, which stands out against its brown body. Its beak is sharp and pointed, suggesting it might be a predator or a scavenger. The bird's eye is large and round, giving it a keen and alert expression. The background is a blur of yellow and brown, indicating that the bird is in a natural environment, possibly a field or a forest. The bird's position in the frame and its direct gaze into the camera suggest that it is the main subject of the video. The video style is realistic, with a focus on the bird's unique features and the natural setting."
+    ],
+    "video_ids": [
+        "rjLrfQDBZfI_10_0to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bird with a vibrant blue crest, brown body, sharp and pointed beak, and large round eyes.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bird that matches the description: it has a vibrant blue crest (visible in close-up shots), a brown body with streaked patterns, a sharp and pointed beak, and large round eyes. The bird's features are clearly depicted, and no elements contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a bird with has the description provided it has a vibrant blue crest,fe in the-up),), a brown body ( as patterns, a sharp and pointed beak, and large round eyes. The bird's features are consistent visible, and the additional contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..90f4113f62c62a2c4bed50be65b1b8082153a91d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1273-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0c446878f7b68de3ef1574e870808df2f26258db11a0443bc516a0bf0bb3955
+size 299797
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..520850599adbb67ebff9af944e3eaf559ed62aa9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two women sitting on a bench, engaged in conversation. The woman on the left is holding a cup of coffee and a cell phone, while the woman on the right is also holding a cell phone. Both women are wearing colorful dresses and headscarves, and they have handbags with them. The bench they are sitting on is blue, and there is a backdrop of a green curtain behind them. The video appears to be a still image, as there is no indication of movement or action. The style of the video is casual and relaxed, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "5OgfsG0SbKc_4_0to116"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two women, one cup of coffee, one cell phone, one cell phone, one dress, one headscarf, one handbag? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two women sitting on a bench. One woman is holding a cup with a straw, and the other is holding a cell phone. Both women are wearing dresses and headscarves. A handbag is visible next to one of the women. All elements mentioned in the description are present, and there are no contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two women sitting on a bench. One woman is holding a cup of a straw, which the other is holding a cell phone. Both women are wearing dresses and headscarves. There handbag is visible on to the of the women. The the mentioned in the condition are present, and there are no contradictions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7472b41934f3aae4db265aa5997e218d153abd4f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e6b619520afe45be5ae791739572e5ede5c5ae683e1e3264e5a9355f5210853e
+size 174205
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..9148ae6172a55a8f2cc3f54b70e55c9dc10ea966
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene day on a river, where several boats are docked. The boats, varying in size and color, are moored along the riverbank, creating a picturesque scene. The calm water reflects the boats and the surrounding trees, adding to the tranquility of the scene. The riverbank is lined with lush green trees, providing a natural backdrop to the boats. The sky above is clear, suggesting a bright and sunny day. The video is shot from the perspective of a boat, giving the viewer a sense of being on the river, surrounded by the beauty of nature and the peacefulness of the water."
+    ],
+    "video_ids": [
+        "yMKJR5n4gLw_48_218to348"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Several boats of different sizes and colors, moored along the riverbank.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple boats of varying sizes and colors moored along the canal bank, matching the description. The boats are clearly visible, and their moored state is consistent with the 'moored along the riverbank' condition. Additional elements like solar panels and surrounding greenery do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows several boats of different sizes and colors,ored along a river,, which the description provided The boats are white visible and and their arrangementored state is evident with the 'moored along the riverbank' condition. The elements like the panels on the greenery do not contradict the description description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..295eb9c75268175e405414eb417a6f935742e64e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:875dee0846e84be8b1e1ca3881529368db88f0e00a06cff02ac2d8fb24c77abd
+size 175172
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..abbbddca9d71eabbefecbea833da061bf5233d90
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a blue shirt and face mask, standing in a room with a blue curtain in the background. He appears to be speaking or gesturing, possibly addressing an audience or a group of people. The style of the video is candid and informal, capturing a moment in a public setting. The man's facial expression and body language suggest that he is engaged in a conversation or presentation. The room has a simple and functional design, with the blue curtain providing a backdrop for the man's activities. The overall atmosphere of the video is one of casual interaction and communication."
+    ],
+    "video_ids": [
+        "6eErcMSa3g4_1_0to198"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue shirt and face mask, appearing to speak or gesture.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue shirt with a face mask hanging around his neck, and he appears to be speaking or gesturing. The background elements, such as other people and blue curtains, do not contradict the core description and are consistent with a public or event setting.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue shirt and a face mask, around his neck, and he appears to be speaking or gesturing with The man includes, such as the people and a curtains, do not contradict the core description and are acceptable with a setting speaking indoor setting.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..32f3183be7751242f95ffac387cedb545bf2bb11
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67b57be99b8d9256ef9cb2e2b85234606ba8af686fe9245ab9745a119ff33d80
+size 167711
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..b85bc3e4c9a3b9c98b4a616d4d6b5602757ea1a3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a drone shot of a luxurious estate with a large pool and a green lawn. The estate is surrounded by a desert landscape with mountains in the background. The pool has a unique design with multiple water features and is surrounded by a well-manicured lawn. The house itself is large and has a grand entrance with a circular driveway. The video is likely meant to showcase the estate's amenities and the surrounding landscape. The style of the video is aerial and it provides a comprehensive view of the property and its surroundings."
+    ],
+    "video_ids": [
+        "e5oDoVn5pyo_6_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Large pool, green lawn, desert landscape, mountains, house with a grand entrance and circular driveway.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a large pool with winding water features, a well-maintained green lawn, a desert landscape with sparse vegetation, distant mountains, and a grand house with a prominent entrance and circular driveway. All core elements described are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a large pool, various edges features, a green-maintained green lawn, a desert landscape with mountains vegetation, mountains mountains, and a house house with a circular entrance and circular driveway. The elements elements of in present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5e5539c9b6f82bdf7a10ccfa60ca87538c934e25
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0544da8d583124bde1941007a55f8b72338151dbe9f45819377d80da75386904
+size 157730
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..c5806b54b0e2ded5ace57d68899e0a6d44b823d8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man driving a car through a rural landscape. The man is wearing a blue sweater and jeans, and he is holding the steering wheel with both hands. The car is a modern SUV with a beige interior. The landscape outside the car is lush and green, with trees and fields visible through the windows. The sky is overcast, suggesting that it might be a cool or cloudy day. The man appears to be focused on the road ahead, indicating that he is driving carefully and attentively. The overall style of the video is realistic and naturalistic, capturing the everyday experience of driving in a rural setting."
+    ],
+    "video_ids": [
+        "VRZ5LEe340M_50_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a modern SUV? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man sitting in the driver's seat of a modern SUV, holding the steering wheel and speaking. The interior and exterior views confirm the vehicle is a modern SUV, and the man is the central subject, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man driving in the driver's seat of a vehicle SUV. which the steering wheel. driving. The interior of exterior of of the presence is a modern SUV, and the man is the primary object of fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f0a53364cc83aa608b1c482b79822a1b31f1cdd6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fcd13e4e391a7fc3639e92269b29fbfc152d7ff3e86b714c1b28e4c1ad4e092d
+size 274655
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..fc1e73aa4bf9842b0fa133a776e11d3163af61a6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a blue Mustang car driving down a street lined with parked cars. The car is in motion, with the driver visible through the windshield. The street is lined with trees and buildings, creating a suburban setting. The car's shiny blue exterior stands out against the more muted colors of the surroundings. The video is shot from a perspective slightly behind the car, giving a sense of movement and direction. The overall style of the video is realistic, with no visible filters or artistic effects. The focus is on the car and its journey down the street, with the background serving as a simple, unobtrusive backdrop."
+    ],
+    "video_ids": [
+        "TkuEOO1m7G4_6_0to134"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue Mustang car, parked cars, trees, buildings? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by prominently featuring a blue Mustang car driving toward the camera. It also includes parked cars on the left side, trees in the background, and buildings visible on the left side of the frame. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting featuring a blue Mustang car, on the camera. The is includes parked cars, the side side of trees lining the background, and buildings further in the right side of the street. The these elements of in present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..39cb4954d8b36f2a281e0939e2261ec1b71ecbd8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34a086d45e8cc21f413cf892ef96b3c9b01548d63ba6858487935dae2b726695
+size 202493
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..562cf0748f1303c3ae1166314b20766f9339c581
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a golf club hitting golf balls in a practice room. The golf club is white and has a black grip. The golf balls are white and have orange, pink, and yellow stickers on them. The golf balls are lined up on the green grass. The golf club is swinging towards the golf balls. The golf balls are in motion. The golf balls are being hit by the golf club. The golf balls are flying through the air. The golf balls are landing on the green grass. The golf balls are rolling on the green grass. The golf balls are in the air. The golf balls are on the green grass. The golf balls are on the ground. The golf balls are on the green grass. The golf balls are on the ground. The golf balls are on the green grass. The golf balls are on the ground. The golf balls are on the green grass. The golf balls are on the ground. The golf balls are on the green grass. The golf balls are on the ground. The golf balls are on the green grass. The golf balls are on the ground. The golf balls are on the green grass. The golf balls are on the ground. The golf balls are on the green grass. The golf balls are on the ground. The golf balls are on the green grass. The golf balls are on the ground. The golf balls are on the green grass. The golf balls are on the ground. The golf balls are on the green grass. The golf"
+    ],
+    "video_ids": [
+        "N0VPEC5L8qM_11_17to167"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Golf club, golf balls? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a golf club (a driver) and multiple golf balls on colorful tees, which fulfills the 'Object(s)' condition. The background elements, such as golf bags and a chalkboard wall, do not contradict the core description and are acceptable as additional context.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a golf club interactinga put) and multiple golf balls on a,es. which arefills the 'Object(s)' condition. The presence and, such as the balls and additional golf line,, do not contradict the core description and are acceptable as additional elements.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b906b1c1b73ca79961a558b78da341925dc193c2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9df1fe28575581d69d49a4212cdf4203b592ede37a94798c165557bff3296c40
+size 259363
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..03ff29c0aab0b7578f4861e12272a2e81df54ea4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man in a red sweater and beige jacket standing next to a white sports car with a black interior. He is pointing at the car, possibly explaining its features or discussing its design. The car is parked in a showroom, and there are other cars and people in the background, suggesting that this is a car show or exhibition. The man appears to be knowledgeable about the car, and his enthusiasm is evident in his gesture. The setting is indoors, and the lighting is bright, highlighting the car's sleek design and shiny exterior. The overall style of the video is informative and engaging, with a focus on the car and the man's commentary."
+    ],
+    "video_ids": [
+        "CGwK821taPI_21_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a red sweater and beige jacket, white sports car with a black interior? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a red sweater and beige jacket standing next to a white sports car with a black interior. The core elements described are accurately represented in the video, even though there are additional background elements like other people and posters.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing a red sweater and a jacket, next to a white sports car with a black interior. The man elements of in present represented in the video, including though there are additional cars cars like other cars and cars,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..57518a449f5a3514c0ec6bf2a760c29ef60dab23
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1274-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a77d163837e41004dd9fa97da00343a59a16dbb5b5efd12a755ee38710985b2
+size 207535
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..58bd1436c1b6b5fdbef80346764d3d45fe4a5d12
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a cartoon character dressed in a safari outfit, complete with a hat and khaki clothing, is seen interacting with a small dinosaur. The character is bending down to pet the dinosaur, which is a vibrant yellow color with brown spots. The scene takes place in a lush, green jungle setting, with towering trees and dense foliage surrounding the character and the dinosaur. The overall style of the video is animated and colorful, with a focus on the interaction between the character and the dinosaur. The character's outfit and the jungle setting suggest an adventure or exploration theme."
+    ],
+    "video_ids": [
+        "62kzrGq2VgY_161_52to228"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A cartoon character in a safari outfit and a small yellow dinosaur with brown spots.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a cartoon character dressed in a safari outfit, interacting with a small yellow dinosaur with brown spots. These elements are clearly visible in the initial frames, fulfilling the 'Object(s)' condition. Additional characters and background elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a cartoon character dressed in a safari outfit, including with a small yellow dinosaur that brown spots. The elements match consistent present and the image frames, fulfilling the 'Object(s)' condition as The elements or elements elements are not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a257669112b417a8d15a8541bf24dd8240523a53
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc8bcc02b2e251e3448cd57b5f7705d8e0f3108468c316fbe7c58dbebbdc2b38
+size 171069
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..99b8cf6ba6014ceac228268a3e8e05a4cc63b8d2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a fantastical creature that resembles a dragon or a mythical beast. The creature is depicted in a highly stylized and artistic manner, with vibrant colors and intricate details. The creature's head is prominently displayed, with a large, open mouth revealing sharp teeth. The creature's body is adorned with scales and feathers, and it has a long, serpentine tail. The background of the video is a warm, glowing sky with clouds, suggesting a magical or otherworldly setting. The overall style of the video is reminiscent of a fantasy or mythical theme, with a focus on the creature's majestic and powerful appearance."
+    ],
+    "video_ids": [
+        "8DqWgRbu5z4_5_0to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A fantastical creature resembling a dragon or mythical beast? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a fantastical creature with a dragon-like appearance, characterized by a serpentine body, ornate head with sharp teeth, and feather-like adornments. These elements align with the description of a mythical beast resembling a dragon, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a fantastical creature that dragon dragon-like appearance, characterized by its multipentine body, multipleate scales, a teeth, and vibranted appendments. The features align with the description of a dragon beast or a dragon. thus the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..885c47c8fc9b2a5d3f8cd15dd30c04c18c6055bc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c27d9b93ff7dd5b923899500844d3880ce58b49a41564f222b5c7659b4fd18c7
+size 228619
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..c9f4e2275977a3a5ba8321453fc5dc0b79daf244
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young woman with long red hair is seen in a snowy setting. She is dressed in a fur coat and a silver necklace, giving her a regal appearance. The woman is looking off to the side, her expression serious and contemplative. The background features a wooden building with a snow-covered roof, suggesting a cold, winter environment. The overall style of the video is realistic, with a focus on the woman's character and the wintry setting."
+    ],
+    "video_ids": [
+        "Q70OR07gKWU_25_0to144"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman with long red hair, dressed in a fur coat and a silver necklace.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman with long red hair, wearing a fur coat and a silver necklace, which matches the description. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a young woman with long red hair, wearing a fur coat and a silver necklace. which align the description provided The setting appears lighting elements in not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9d6b6f4cf6f30ce27bbcb2e62bbff0713908f074
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:174b231fd8268f9d398f14ae1cdaf4d7fda109e8d84631061a10714acb82c458
+size 153485
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..e153e828499ea2dadc62fe92d32be240532d40fe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a blue sports car parked on a concrete surface, with a construction site in the background. The car is equipped with large black wheels and a black spoiler on the trunk. The car is stationary in the first frame, and in the second frame, it begins to move forward. In the third frame, the car is in motion, driving away from the camera. The style of the video is dynamic and action-oriented, capturing the car's movement and the industrial setting."
+    ],
+    "video_ids": [
+        "55l0zSrT1Ts_5_52to246"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a blue sports car, specifically an Audi coupe, which matches the description. The car's color and sporty design are clearly visible, and while there are additional background elements (like construction equipment), they do not contradict the core description of a blue sports car.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a blue sports car, which a A R, which matches the description of The car is design, designy design are consistent visible, and the there are additional elements elements likea a and and they do not detr the core description of the blue sports car.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7a8ccfab7f0d214dc2464344cc4a0a22cbf39976
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ffebb2a911bc167ad74563a829546b5497dd4335a49fa410fb24d3e8d6c5fb99
+size 255621
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..9b4b777a95160a6e0fb14871a0785526125d4841
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen holding a small toy train on a wooden table. The train is blue and yellow, and it appears to be a model train. The person is holding the train with their thumb and index finger, and they are looking at it closely. The table is brown and wooden, and there is a small plastic bag on the table as well. The person is wearing a gray sweater, and they seem to be in a room with a white wall in the background. The video captures the person's attention to detail and their interest in the toy train."
+    ],
+    "video_ids": [
+        "2XA_Cw_0ESU_16_117to295"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a small toy train, a brown wooden table, a small plastic bag? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person holding a small toy train on a brown wooden table. A small plastic bag is visible in the background, partially out of focus, which aligns with the description. All core elements are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a person interacting a small toy train, a brown wooden table. There small plastic bag is also in the background, which wrapped of focus. which aligns with the description. The the elements are present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..172660c0bd274ec085548c295d121ba23230d6d4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e829af9e37838fb76e87e7fd844435cdecadca65cad8ca46fada1960a04c0c0d
+size 166390
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..e67a56fb03c23621f6536d28b6def6dfd75c3ba3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man named Peter McLaughlin, a resident, who is being interviewed on a news channel. He is wearing glasses and a blue shirt. The background of the video shows a brick wall with a red cross sign. The news channel is tracking Dorian, a storm, and the man is likely discussing his experiences or concerns related to the storm. The style of the video is a standard news interview, with the interviewee in the foreground and the news ticker and logo in the background."
+    ],
+    "video_ids": [
+        "aI41ibgz2LQ_13_0to157"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man (Peter McLaughlin), glasses, blue shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man identified as Peter McLaughlin, wearing glasses and a blue shirt, which matches the specified 'Object(s)' condition. Additional elements such as the brick wall, Red Cross sign, and news graphics do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing as Peter McLaughlin, wearing glasses and a blue shirt. which align the description 'Object(s)' condition. The elements such as the background wall background red Cross symbol, and text ticker do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..54ca15f1b1f5462dee211bea37523afba4c9a42b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4f4b0b58f3d4e311053a5b1a9e599ad9573f9ab0e463891313c72691e5670b7e
+size 112032
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..772342216072c9bbebba0d595d4b90ac57705963
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit and tie, standing in front of a backdrop that includes a city skyline and the European Union flag. The man appears to be speaking or presenting, as suggested by his open mouth and gesturing hand. The style of the video is professional and formal, likely intended for a news or business broadcast. The man's attire and the backdrop suggest a serious and authoritative tone. The backdrop is colorful and dynamic, with the city skyline and the European Union flag adding a sense of location and context to the video. The overall impression is of a well-produced and polished video, designed to convey information or news in a clear and professional manner."
+    ],
+    "video_ids": [
+        "pzj6MbDVaXk_3_25to157"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, gesturing and possibly speaking or presenting.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed in a suit and tie, standing in front of a backdrop with European Union flags and an Apple logo. He appears to be gesturing with his hands and his mouth is moving, suggesting he is speaking or presenting. These elements align with the described 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man in in a suit and tie, standing in front of a background that a Union and and a urban logo, He appears to be gesturing with his hands, possibly mouth is open, suggesting he is speaking or presenting. The elements align with the description 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..84b0b5f7c0d2636f688eb77c0a839e3b2f22070e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb0027c0e2fb219d27ce5b23f67619a9ac9e016a155a91a01ddd12c99a3d1f0f
+size 118613
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..1be21503d2d7a6e095d0b07b7244bf24847ece28
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a lime green sports car being worked on in a garage. The car is lifted onto a blue hydraulic lift, and the hood is open, revealing the engine. The garage is well-lit, with various tools and equipment visible in the background. The car is the main focus of the video, and the garage setting suggests that it is being serviced or repaired. The style of the video is straightforward and documentary-like, with no additional embellishments or artistic effects. The video captures the process of working on a sports car in a professional setting."
+    ],
+    "video_ids": [
+        "nJLzFli61eM_23_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A lime green sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a lime green sports car elevated on a blue car lift in a garage setting, matching the core description. Additional elements like onlookers and workshop equipment do not contradict the presence of the specified object.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a lime green sports car, on a blue hydraulic lift. a garage setting. which the description description of The elements like thelookers and tools tools are not contradict the main of the lime object.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a19b93bba595624d0bda601103018d511c964bf5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1275-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a53b5435c8c519b41df801d5beedf2d2216908da688ada753d404d9b32c92b97
+size 195847
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..7b86df6211f6a1e677b7febe16858afb795b6acf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a vibrant scene at a car show. The main focus is a striking green sports car, which is parked on a concrete surface. The car is adorned with a black stripe and a black roof, adding to its sleek design. The car's headlights are on, casting a green glow that illuminates the surroundings.  In the background, there are other cars on display, each with their own unique designs and colors. The cars are parked in a row, creating a visually appealing line-up.  There are also people present in the scene, walking around and admiring the cars. Their presence adds a lively atmosphere to the car show.  The overall style of the video is dynamic and colorful, capturing the excitement and energy of the car show. The green sports car, with its vibrant color and unique design, stands out as the star of the show."
+    ],
+    "video_ids": [
+        "T3Wmh8BWSs0_9_0to154"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A striking green sports car, other cars on display, people walking around.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a striking green sports car with custom decals and lighting, which matches the description. In the background, other cars are visible, including a black car with the number '55' on it. People can also be seen standing around the cars, consistent with the 'people walking around' part of the description. No elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a striking green sports car, a black, a, which align the ' of Additionally the background, there cars are visible, indicating a blue and and a door 1'' and it, Additionally are be be seen walking and, cars, which with the descriptionpeople walking around' part of the description. The elements contradict the given description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..313769ef2a5c017c30076e7141aa87dc642fbda4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e00cbfd06b638f162321fb5504bdc08fc7824619b775dfa738ae1d243b434ff
+size 193096
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..334d68726612d13877ebf32a20d6faa67029b729
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with blonde hair, wearing a red top and a necklace. She is looking directly at the camera with a slight smile on her face. The background is a plain, light blue color, providing a simple and uncluttered backdrop that focuses attention on the woman. The style of the video is straightforward and professional, likely intended for a corporate or promotional purpose. The woman's attire and demeanor suggest she may be a business professional or public figure. The lighting is even, and the focus is sharp, indicating a well-produced video."
+    ],
+    "video_ids": [
+        "0XXN9AjAYFs_6_132to330"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with blonde hair, wearing a red top and a necklace, looking directly at the camera with a slight smile.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair, wearing a red top and a necklace, and she is looking directly at the camera with a slight smile. These elements match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair, wearing a red top and a necklace. looking she is looking directly at the camera with a slight smile. The elements match the description provided, the questionObject(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..74cccd8051c4c60ce280f51c3d9f2b544118df84
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c61d0ea1068f064edd7589d20ff3320b5449635080df8ee9be8fbe38bbc4cbe5
+size 101651
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..b91201b347fc408be2da940a2c0b5a0394da5390
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic scene of a dolphin leaping out of the ocean. The dolphin is mid-air, its body arched gracefully as it propels itself upwards. The water around it is splashing vigorously, creating a misty spray that adds to the sense of motion and energy. The dolphin's sleek form is highlighted against the vibrant blue backdrop of the sea, which stretches endlessly into the horizon. The camera follows the dolphin's movement, maintaining focus on the animal as it ascends higher into the air. The sequence showcases the dolphin's powerful leap and the dramatic splash it creates upon re-entry into the water. The overall scene is one of natural beauty and the raw power of marine life."
+    ],
+    "video_ids": [
+        "3100391f71123f1465d01e405ba859ebbfde8122ce9b9e93fa764c4704e6ae95"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A dolphin, water splashes, and mist.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a dolphin leaping out of the water, accompanied by significant water splashes and mist, which aligns perfectly with the described conditions. The dolphin's dynamic movement and the resulting spray of water are vividly captured, fulfilling the 'Object(s)' requirement without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a dolphin inaping out of the water, creating by significant water splashes and mist. which ares with with the ' '. The dolphin is movement movement and the resulting water of water are thely captured, fulfilling the 'Object(s)' condition.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8143d89ceb74d94810eade38820f8d1518255e31
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:66c04045eb8cde4e2338216126778ada368389eb03d4ac835c7cd645baed580f
+size 273508
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..20d52d9c9d31904ed1ed6706e8cabd6ea6731dcb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene underwater scene featuring a polar bear in a rocky environment. The polar bear is partially submerged in clear water, with its head and upper body visible above the surface. The bear appears calm and relaxed, occasionally moving its head slightly. The background consists of large, textured rocks that create a naturalistic setting for the bear. The water is crystal clear, allowing for a detailed view of the bear's fur and the surrounding environment. The lighting is soft, suggesting a daytime setting with natural light filtering through the water. The bear's reflection can be seen on the water's surface, adding depth to the scene. The overall atmosphere is peaceful, highlighting the bear's natural habitat and behavior in a controlled environment."
+    ],
+    "video_ids": [
+        "ddf88f8fef5cf362395f085395cc0bc85ff0d1e8c12b17487a1fc2fcaa8f5c81"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A polar bear? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a polar bear interacting with its environment, both above and below water, which matches the 'Object(s)' condition of a polar bear. The bear's physical features and behavior are consistent with those of a polar bear, and there are no elements that contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a polar bear, with its environment, which in and below the. which align the descriptionObject(s)' condition of a polar bear. The presence's white characteristics, the are consistent with what of a polar bear, and there are no elements in contradict this description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e5749f32114d4703ca809556251f614e5c1b0f66
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e4b09805dd260e0677a2eab016c08e2b10e278c4c1adbf05e8ba83769c24121
+size 262285
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..9c6ce189cb492a956e99634108fb6bf32fa23168
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic scene of a blue and white speedboat cutting through a calm lake. The boat, with its sleek design and vibrant colors, is the main focus of the video. It's moving at a high speed, leaving a trail of white water behind it, indicating its swift movement. The lake itself is vast and serene, with a clear blue sky above and lush green trees lining the shore in the background. The boat's reflection is visible on the water's surface, adding a sense of symmetry to the scene. The overall style of the video is dynamic and energetic, capturing the thrill of speedboat riding in a beautiful natural setting."
+    ],
+    "video_ids": [
+        "qVsWoV25h-A_0_138to262"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue and white speedboat? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a speedboat with a blue and white color scheme, matching the description. The boat is in motion on water, and its design and colors are consistent with the 'blue and white speedboat' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a blueboat with a blue and white color scheme. which the description. The boat is in motion on a, and the design and color are consistent with the 'Object and white speedboat' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f67cd54770a4392b6c6426b97d42681e817c5e0b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6bf224487de45a6551875b48f3e2f56d4e800ee6b3cd7327ed8b291f516e5206
+size 191830
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..144ca4d534fba3d5edd418ffc0a45362e9b6f067
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of two birds in flight against a blurred natural background. The birds, likely doves or pigeons, are shown in mid-flight, their wings spread wide as they soar through the air. The focus is primarily on one bird, which dominates the frame, while the second bird appears slightly out of focus in the background. Both birds exhibit a mix of white and gray feathers, with darker markings visible on their wings and tails. The lighting suggests it might be late afternoon or early evening, casting a soft glow on the birds' feathers. The camera remains steady throughout the sequence, maintaining a consistent angle that emphasizes the birds' graceful movement and the fluidity of their flight. The overall scene conveys a sense of freedom and tranquility, highlighting the beauty of nature in motion."
+    ],
+    "video_ids": [
+        "d08953fb0bb912c66225a2d1ba39eb1abf822096b0ebddbd296425555d5488bb"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two birds, likely doves or pigeons.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two birds in flight, with visible wing movements and body shapes consistent with doves or pigeons. Their plumage appears to be primarily gray and white, which matches typical dove/pigeon coloring. The motion and context strongly support the description without contradicting it.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two birds in flight, which one wings patterns and a shapes that with doves or pigeons. The plumage, to be predominantly white and white, which is the dove origeon color. The setting and the suggest suggest the identification of anying it.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cda53d90379400a3688e6dbb5d5d867352ada0e6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d0d9928d0604e41817d158f3ab74ee52fd8283256ee068545ccf0201778856f
+size 197843
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..a813eadf2904acfc0ed5bdf2f200c22ff04b3654
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in front of a window, holding a ladder. He is dressed in a white t-shirt and appears to be speaking or presenting. The window behind him is brightly lit, suggesting it is daytime. The man's expression is serious, and he seems to be focused on his task or message. The ladder he is holding is white and appears to be sturdy. The overall style of the video is straightforward and professional, with a focus on the man and his message."
+    ],
+    "video_ids": [
+        "1hFBSgnvbj4_13_0to123"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man, window, ladder? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man standing in front of a window with a ladder beside him. All three specified objects are present and accurately depicted in the scene without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man holding in front of a window, a ladder. him. The the elements elements ( present: match depicted. the scene. any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b24d5b7be7cd5f21dabdf0aae8ab962ff738dc3a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5684774bb9837f05ed1b260615266328edff22f2f485518cc496dfcd66ca631f
+size 134349
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..ad74e5e49d76c5fed1f153c327d49d2271c27618
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young woman is seen sitting at a desk, surrounded by a colorful array of toys and stuffed animals. She is wearing a red shirt and has a warm smile on her face. In her hands, she holds a basket filled with various toys, including a teddy bear and a toy car. The desk is cluttered with more toys and a cat is seen sitting on the desk, adding to the playful atmosphere. The room is filled with a sense of joy and creativity, as the woman seems to be enjoying her time with the toys. The video captures a moment of simple happiness and playfulness."
+    ],
+    "video_ids": [
+        "qWVFyFd259c_8_159to305"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman, a basket filled with various toys (including a teddy bear and a toy car), a desk, more toys, and a cat.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman sitting at a desk holding a basket filled with various toys, including what appear to be plush toys resembling a teddy bear and other figures. A cat is visible in the background, resting on a drawer. The desk and surrounding area are cluttered with more toys, matching the description. The video does not contradict any core elements of the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman interacting at a desk with a basket filled with various toys, including a appears to be a toys and a teddy bear and a items. The toy is also on the background, and on the surface. The desk is surrounding area are filleded with more toys, which the description. The presence successfully not contradict the of elements of the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..09a839ba47b2df13479e76e9fe1567ae0f8b4eb9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1276-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:47dfbb0029603a49cde2b35ffe08ff520be3f329b4c2b1252a70aa0b898eb082
+size 196034
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..53ab6abb0c51d944ba8b46a82bbc856a8165d6a8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the vibrant colors of autumn leaves as they change and fall from the trees. The leaves are scattered across a concrete surface, creating a beautiful mosaic of red, yellow, and green hues. The leaves are in various stages of decay, with some still retaining their vibrant colors, while others have turned brown and withered. The leaves are not neatly arranged but rather randomly scattered, adding to the natural beauty of the scene. The concrete surface provides a stark contrast to the organic shapes and colors of the leaves. The video is a celebration of the changing seasons and the beauty of nature."
+    ],
+    "video_ids": [
+        "L8RNMpObGaQ_9_0to161"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Autumn leaves? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a collection of autumn leaves in various colors such as red, yellow, green, and brown, scattered on a textured surface. These leaves are consistent with the description of 'Autumn leaves,' and no elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a collection of leaves leaves in various colors such as red, yellow, orange, and brown, which on a blue blue. The elements are typical with the typical of autumnAutumn leaves,' which there additional contradict this description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..70fefd69970e41986be102909907e84a6470dc8f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17cd42ac5e56d549b1e6859c9df15ff6cf5987b97fd50c28806080687faa3e2c
+size 202437
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..708c3ccc2d2f129d187b9e978ca48ee2d6da9b9c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the thrilling experience of a boat ride on choppy waters. The viewer's perspective is from the deck of the boat, looking out towards the horizon. The boat is white and features a wooden deck, adding a touch of elegance to the adventure. The water around the boat is a mix of white and blue, indicating the roughness of the sea. The sky above is filled with dark clouds, suggesting an impending storm. The overall style of the video is dynamic and adventurous, capturing the essence of a thrilling boat ride on the open sea."
+    ],
+    "video_ids": [
+        "5U6U4q0jUdc_21_0to128"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Boat (white with wooden deck), water (choppy, white and blue), horizon, dark clouds in the sky? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white boat with a wooden deck, choppy water with white and blue hues, a visible horizon, and dark clouds in the sky, all of which align with the described conditions. The presence of the 'The Boat Show' logo does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a white boat with a wooden deck, whichpy water with white and blue tones, and visible horizon, and dark clouds in the sky. all of which align with the description '. The presence of the boathor horizon'' logo does not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..01a10a00475d20744e06ef555bc8dc19121e9a09
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9d7878d61ac2e0b8578631fcebd9d60c058f781c8197eba61f3ce978eace5e0f
+size 189169
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..b12711bad3f0bf0498ef362e16b17e214405e6bd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic scene on a highway. Two sports cars, one yellow and one gray, are driving side by side. The yellow car is in the left lane, while the gray car is in the right lane. They are both moving at a high speed, as indicated by the blurred background. The highway is surrounded by a lush green landscape, with trees and grass visible on both sides. The sky above is clear and blue, suggesting a sunny day. The overall style of the video is dynamic and fast-paced, capturing the thrill of driving at high speeds."
+    ],
+    "video_ids": [
+        "X7aDv6YhCdM_51_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two sports cars, one yellow and one gray.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two sports cars driving side by side on a highway: one is yellow and the other is gray. Both vehicles are identifiable as sports cars, matching the description. Additional elements like other cars and roadside infrastructure do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two sports cars, on by side on a road. one is yellow and the other is gray. The cars have sleek as sports cars based and the description provided The elements such the cars, the scenery are not contradict the core description but}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7ef403583aa3341a30f3b9d86ac20d9a484d0729
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:faec313115520a64ce802c475757ea236e14fb325330f8e4c00bc4d6dc8b5bcc
+size 220940
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..726d547c0eee07c1671bdd8f648e4cdd6fb1000b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures an underwater scene teeming with marine life and vibrant coral formations. The camera glides smoothly over a lush coral reef, showcasing a variety of corals in different shapes and sizes, ranging from branching to rounded forms. The colors of the corals vary from deep greens to lighter hues, creating a rich tapestry beneath the water's surface.  As the camera moves forward, it reveals a school of fish swimming gracefully among the corals. These fish, characterized by their dark bodies with white stripes, glide effortlessly through the water, adding a dynamic element to the serene underwater landscape. The fish appear to be exploring the reef, occasionally darting between the branches and crevices of the coral formations.  The camera continues its journey, capturing the intricate details of the coral structures and the fluid motion of the fish. The interplay of light and shadow enhances the visual appeal, highlighting the textures and contours of both the corals and the fish. The overall atmosphere is one of tranquility and natural"
+    ],
+    "video_ids": [
+        "1e621229ce3128cb4de65fdb597128e0176a92eab82ffed9af13bd16bcf49fbc"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Corals, fish? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows corals and fish as requested. The underwater environment is filled with various types of coral formations, and a fish is clearly visible swimming among them. The presence of other marine life and coral structures does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful aals and fish, described. The image scene is depicted with various cor of cor,, which there school can visible visible swimming among them. The presence of additional marine life and the types does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cf6bde109ca448cff9ed351cea98fc892edaa7ba
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4f19cf37919c43bd2929649db034710d6c7aae35b9a4810a327266e3102bf6ff
+size 258108
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..c9a074fab1ee0c2523fae11678ea6104e5c950a8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a large black lizard with yellow spots, possibly a monitor lizard, as it moves through a lush green forest. The lizard is seen climbing a tree, its long tail and powerful limbs gripping the branches. The forest is dense with green foliage, and the lizard's dark color contrasts sharply with the surrounding greenery. The lizard's movements are slow and deliberate, suggesting a careful and methodical approach to its surroundings. The video is a close-up shot, focusing on the lizard and its immediate environment, and the camera follows the lizard as it ascends the tree. The overall style of the video is naturalistic, capturing the lizard in its natural habitat and showcasing its unique features and behavior."
+    ],
+    "video_ids": [
+        "KYFNKlp9wks_13_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large black lizard with yellow spots (monitor lizard).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large lizard with a predominantly black body covered in yellow spots, which matches the description of a monitor lizard. The lizard is resting on branches in a natural, leafy environment, and its physical features (size, coloration, and pattern) are consistent with a monitor lizard, specifically resembling a water monitor or similar species. There are no elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a large black with a predominantly black body and in bright spots, which matches the description of a monitor lizard. The lizard is per on a, a natural setting outdoory environment, which the appearance characteristics,color, coloration, and spots) align consistent with the monitor lizard. fulfilling one a species monitor or a species. The are no additional in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..98c750be12b42550801b1f4b585848543d3b8af0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fa28f4ca971ae6cf65babe17d5a98dd39c76b6b665548dd57181af8d08e8948a
+size 167325
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..1a2057d323a3554e3671c35fcab24a524956bf73
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a gray Jeep Wrangler driving on a road. The Jeep is equipped with a black roof rack and has a spare tire mounted on the back. The vehicle is in motion, and the camera follows it from a front three-quarter angle. The road appears to be paved, and the surrounding environment is not clearly visible. The style of the video is a straightforward, real-world depiction of the vehicle in motion, with no additional embellishments or artistic effects."
+    ],
+    "video_ids": [
+        "XhKzxNwKlTQ_21_0to130"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A gray Jeep Wrangler with a black roof rack and a spare tire mounted on the back.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a gray Jeep Wrangler with a black roof rack, which matches the description. Although a spare tire is not clearly visible on the back, the presence of a roof rack and the overall vehicle appearance align with the core description. The video does not contradict the primary elements specified.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a gray Jeep Wrangler with a black roof rack, which matches the description. The the spare tire is not visible visible in the back of the overall of a roof rack is the overall design design align with the given description. The video does not contradict the given elements mentioned in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1d885b2d41a7e8911aa6b741b270b6b5e144e832
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6a233bd9f120eb499bdc7650e272b06eaba659c6bad17c6d9a282ce8d66884d0
+size 202032
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..665dd702dca854f234272d68499ee459c8673df1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a car engine, focusing on the intricate details of the engine components. The style of the video is a time-lapse or a series of still images, capturing the engine from different angles and perspectives. The engine appears to be well-maintained and clean, with various hoses, belts, and other parts visible. The video does not contain any text or additional elements, focusing solely on the engine. The lighting in the video is bright, highlighting the metallic surfaces and the complexity of the engine's design. The video is likely intended for educational or promotional purposes, showcasing the engineering and design of the car's engine."
+    ],
+    "video_ids": [
+        "uiTX4dVHzAc_60_0to160"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Engine components including hoses, belts, and other parts.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows various engine components such as hoses, belts, and other mechanical parts within the open engine bay of a car. The focus remains on these components throughout the sequence, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a engine components such as hoses, belts, and other parts parts, the engine engine bay of a car. The presence is on the elements, the frames, fulfilling the 'Object(s)' condition as}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ffe6cde75ca427b4e3deec2880720ca926646a4e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:beadde85fe6850f39d0cc8eddedcef5b37c11d487c51b937d461a3431842e3c7
+size 114829
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..f2a237d26d8eff225b72eea3e24370c45272c9b3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, three young women are gathered around a table, each holding a cupcake. The first woman, with her eyes closed, is smiling as she holds her cupcake. The second woman, wearing a white shirt, is leaning on the table with her chin resting on her hand, also holding a cupcake. The third woman, wearing a blue and white plaid shirt, is looking at the first woman with a smile on her face. The table is adorned with a pink potted plant and a blue vase. The scene is set in a kitchen with a yellow wall and a white shelf in the background. The overall atmosphere of the video is cheerful and light-hearted."
+    ],
+    "video_ids": [
+        "G6t10UJckqE_14_0to178"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three young women, a pink potted plant, a blue vase, a table, a cupcake (3), a white shirt, a blue and white plaid shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three young women at a table, one wearing a white shirt and another a blue and white plaid shirt. A pink potted plant is visible on the table, and there are three cupcakes (two being held, one on a plate). A blue vase is also visible in the background on a shelf. All specified objects are present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts three young women sitting a table, each of a white shirt and the wearing blue and white plaid shirt. There pink potted plant and visible on the table, and there are cupcakes cupcakes,one with held by one on the plate). The blue vase with also present in the background. a shelf. The elements elements are present, match the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8f385ddd3ec06b29f1014cf95296ad9a7ae76c60
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1277-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eedbd72681293c7cc345ff4ef358fbe9cd51cd3bdf2a4140302f202c81e517dc
+size 142796
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..21dd8c367279c397e7bee008afa79b5227f31192
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a wide-brimmed hat and sunglasses, standing in a grassy field with mountains in the background. He appears to be speaking or singing, as his mouth is open. The man is wearing a dark-colored shirt and has a pair of headphones around his neck. The style of the video is casual and outdoor, with a focus on the man and his surroundings. The lighting suggests it is daytime, and the overall atmosphere is relaxed and leisurely."
+    ],
+    "video_ids": [
+        "b9ofqbs8Pco_58_54to195"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a wide-brimmed hat, sunglasses, dark-colored shirt, headphones around his neck? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a wide-brimmed hat, sunglasses, a dark-colored shirt, and headphones around his neck. These elements are clearly visible and consistent with the description, even though the background and some reflections may suggest AI generation. No conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person wearing a wide-brimmed hat, sunglasses, and dark-colored shirt, and headphones around his neck. The elements match consistent visible and match with the description provided indicating though the background is the details in not a generation.\"\n The contradictions elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3bd75f48ff3e49f02650e6fbea32fcc9da46d7b3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83369c1c3dd526e27f4b5a4db2513b8b2f4053c8de84db55c89a9554a7d10a65
+size 192193
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..444ba9ec8dcefd9d6d0e5d3a3d1504518a12e52b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a football player in action, wearing an orange and blue uniform with the number 18. The player is seen in three different frames, each showing him in a different position on the field. In the first frame, the player is seen running towards the camera, his body leaning forward in anticipation. In the second frame, he is seen in mid-air, having just jumped to catch a pass. In the third frame, he is seen running towards the end zone, his arms outstretched in celebration. The video is shot from a low angle, emphasizing the player's athleticism and the intensity of the game. The background is filled with the crowd, adding to the excitement of the moment."
+    ],
+    "video_ids": [
+        "LCfZxdZif6E_5_0to133"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A football player in an orange and blue uniform with the number 18.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing an orange and blue uniform with the number 18 clearly visible on the jersey. The player is also wearing a helmet with the Denver Broncos logo, consistent with the team's colors and uniform design. The background, though blurred, appears to be a stadium setting, which is appropriate for a football game. There are no elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing an orange and blue uniform with the number 18. visible on the back. The player is seen seen a blue, a number Broncos logo, which with the team's colors and branding.. The player suggests including not, appears to be a football setting, which is typical for a football game. The are no elements in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b44912d7d1abed1f6ece5fb981cc44718203cfbb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b25b33155a3c343bda19f168744de9afd0a852ade2a46470aa43f8ac4dc612e2
+size 370468
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..4dd31667602527712965a004bb5b8554a1152cc9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene underwater scene featuring a polar bear in a zoo enclosure. The bear is swimming gracefully near the surface of the water, its head and part of its body visible above the waterline. The environment around the bear includes large, rocky formations that mimic a natural habitat, providing a textured backdrop to the scene. Overhead, a white canopy structure is supported by metal poles, offering shade and protection from the sun. The water is clear, allowing for a detailed view of the bear's movements and the surrounding rocks. The bear appears calm and relaxed, moving slowly through the water, occasionally dipping its head below the surface. The overall atmosphere is tranquil, highlighting the bear's natural behavior in a controlled setting."
+    ],
+    "video_ids": [
+        "cf250f8bce224874a1893a24f7b297dfd0aa6eea6a85ce3b5658650b24a30e91"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A polar bear? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a polar bear swimming underwater in a pool with rocky surroundings. The bear's white fur, body shape, and swimming behavior are consistent with a polar bear. The scene is split between above and below water, but the polar bear remains the central and identifiable subject throughout.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a polar bear, in, a pool, a textures. The bear is white fur and the shape, and the posture are consistent with the polar bear. The presence is AI between the and below water, but the presence bear is the central focus identifiable object.\"\n the}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4ab7d2f65d593b2832b15ba70a9b2af15b6d8459
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff97163392e4f2ecf665592c0866f4fc4af356c76fcb6cad3298912e1f1f648c
+size 411287
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..c27a0ea50fd937dd69a5053cf5d4a7cc0f10e0d2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young child is seen sitting in a car seat, which is pink and decorated with stars. The child is wearing a gray shirt and has a blue pacifier in their mouth. The car seat is positioned in the back seat of a car, and the child appears to be looking out the window. The car is moving, as indicated by the blurred background. The child seems to be in a relaxed state, possibly enjoying the ride. The overall style of the video is casual and candid, capturing a moment of a child's life during a car ride."
+    ],
+    "video_ids": [
+        "0SPv1uY0Yvc_30_266to466"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young child in a car seat, a pink car seat, and a car.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a young child seated in a pink car seat inside a car. The child is secured with a harness, and the car seat features a princess-themed design. All core elements described in the condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a young child seated in a car car seat, a car. The child is wearing with a seat, and there car seat has a pattern-themed design with The elements elements of in the condition are present, accurately depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0e7e6eb0d823921b2a8252b94c61e23a6f3950d4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3788a29d2ec94958660238c62a1c8896b58ff23232448f4caee3ffa96b94e5da
+size 183234
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..07470c05459c8509aac4cf7c91123c0a9bfe2a26
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a Barbie doll in a blue dress with a star pattern, standing in a room with a pink door and a window. The doll is holding a small purse. The room has a pink and white color scheme, with a bookshelf and a television in the background. The doll is positioned in front of the window, and the pink door is closed. The overall style of the video is playful and colorful, with a focus on the Barbie doll and her surroundings."
+    ],
+    "video_ids": [
+        "8l1F3h_BF7A_19_0to167"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Barbie doll in a blue dress with a star pattern, holding a small purse? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a Barbie doll wearing a blue dress adorned with a star pattern and holding a small purse. The doll is the central focus, and all key elements of the description are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a Barbie doll wearing a blue dress with with a star pattern. holding a small purse. The setting is positioned central focus, and the elements elements of the description are present represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6796f2661cd1f658eba6e6a9bedf2f561f0d71cd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f0c2924fe1b9a329b7742a94935b7a4478dba3e0e3bb311c315f8a4df4e143d0
+size 159046
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..186624d2d5e66d4f1da8e2b455baf8fbb9d77bb4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a small, animated figurine of a princess with brown hair and large, expressive eyes. She is dressed in a yellow dress with a puffy skirt and a matching yellow headband. The princess is standing on a white surface, and her pose changes slightly between frames, giving the impression of movement. The style of the video is simplistic and cartoonish, with a focus on the princess and her attire. The background is plain and does not distract from the main subject. The video is likely aimed at a younger audience, given the style and subject matter."
+    ],
+    "video_ids": [
+        "Y8ET5ivuA1c_7_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small, animated figurine of a princess? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a small, stylized figurine of a princess, specifically Belle from Disney's \\\"Beauty and the Beast,\\\" dressed in her iconic yellow gown. The figurine is animated in the sense that it is shown moving slightly as the camera pans down, which aligns with the 'animated' aspect of the description. There are no conflicting elements; the focus remains on the princess figurine.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a small, animatedized figurine that a character, characterized resembling from Disney, 'Beauty and the Beast.\\\" as in a iconic yellow gown and The figurine has animated, a sense that it is shown from slightly, if camera pans around, which iss with the 'animated' aspect of the description. The are no additional elements in the focus remains on the figur figurine,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4e8c539304f6dbc16cbdedaf67997d12e771c4cb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:82c5968229145b261c9f2a1055323c8e8832709e8f57010ac32baf3e33e9c139
+size 201663
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..7e62aee3c9a75316ace6ea2f78d3e1f9f8c469e9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a soccer player in action, wearing a green jersey with the number 11 and a Nike logo. The player is seen in three different positions, showcasing his agility and skill on the field. The background is a blur of spectators, indicating that the game is taking place in a stadium filled with fans. The player's focused expression and dynamic poses suggest a high-stakes match, with the player fully engaged in the game. The video is a dynamic and engaging portrayal of a soccer player in action."
+    ],
+    "video_ids": [
+        "qPL5h-yt8RI_27_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A soccer player in a green jersey with the number 11 and Nike logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a green jersey with the number 11 and a Nike logo, which matches the core description. The background is blurred, but this does not contradict the object condition. The player's attire and visible markings are consistent with the specified requirements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a soccer player wearing a green jersey with the number 11 and a Nike logo. which matches the description description. The player and a, but the does not contradict the description condition as The player's attire and number number are consistent with the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3028bf104d1ca882e27b0368e6a2493b681a8e26
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2db87d4fce326b9c8a19240f2200664c108e24720712f4d35628768e7eaf8f50
+size 237440
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..78a03add836511c7e26e6c15508538d98ddca00a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman and a man are engaged in a conversation. The woman, who is the main subject, is wearing a floral print blouse and is gesturing with her hands as she speaks. The man, who is seated across from her, is wearing glasses and is attentively listening to her. They are both seated at a dining table, which is covered with a white tablecloth. The room they are in has a white door and a wall with a green plant painted on it. The overall style of the video is casual and intimate, capturing a moment of conversation between two people."
+    ],
+    "video_ids": [
+        "c5e4jaRlGBU_43_0to135"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman and a man? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman and a man engaged in conversation, which fulfills the 'Object(s)' condition. The woman is visible in full, and the man is partially visible from the side, consistent with the description. No elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a woman and a man sitting in a at which fulfills the 'Object(s)' condition. The presence is wearing on the, and the man is partially visible, the side, indicating with the description of The additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6a53f5d31ed233cafe8d437a0fe52c656f58e44f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1278-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90d04237967e758c5a8ea59112e23debd728bf4f9d2e07f1ef1a0861d9941aba
+size 199725
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..b287c6ae8685636c9d45bafbfe6609a94cbcc6a4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a hockey game. Two players, one in an orange and blue uniform and the other in a white and black uniform, are engaged in a fierce competition on the ice. The player in the orange and blue uniform is in the process of taking a shot at the goal, his body leaning forward with intensity. The player in the white and black uniform is attempting to block the shot, his body positioned defensively. The scene is set against the backdrop of a hockey rink, with the goal and the spectators visible in the distance. The video is a freeze-frame of a high-energy moment, capturing the tension and excitement of the game."
+    ],
+    "video_ids": [
+        "3kgLg3htjuo_6_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two players (one in orange and blue, one in white and black), a goal, spectators? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two hockey players: one in orange and blue (Edmonton Oilers) and one in white and black (Anaheim Ducks), as described. Spectators are visible in the background, and while a goal is not directly visible, the context of a hockey game implies its presence. The core description is accurately fulfilled.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts two players players, one in an and blue,numbermonton Oilers) and one in white and black (Losheim Ducks), positioned well. Thereators are not in the background, and there the goal is not explicitly visible, the players of the hockey game implies the presence. The image elements is largely represented.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c282689728cd3cd92ee8bc4c519e4570e2ee318f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:27da6a3b83702e060a096d9e3334be7f3c0c6acafce308362969861618d492ac
+size 249070
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..a109dc497ec0f9d3d77a0098725a40e186fdcd71
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with short brown hair, wearing a pink blouse and a necklace. She is smiling and looking directly at the camera. The background is blurred, but it appears to be an outdoor setting with trees and sunlight. The style of the video is a simple, straightforward portrait with a focus on the woman's face and expression. The lighting is bright and even, highlighting the woman's features without creating harsh shadows. The overall impression is one of warmth and friendliness."
+    ],
+    "video_ids": [
+        "ZZG5i4Kd0iw_4_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Woman with short brown hair, wearing a pink blouse and a necklace, smiling and looking directly at the camera? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with short brown hair, wearing a pink blouse and a necklace, smiling and looking directly at the camera. These elements match the description exactly, and there are no conflicting elements present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with short brown hair, wearing a pink blouse and a necklace. who and looking directly at the camera. The elements match the description provided, indicating there are no additional additional present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6346bfcb1864a865fbfb9abaea798f9fad293151
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6dc0a497db79e7728bea62d186566a1e9fdcca9ff9187a2dcce5755d3641e1c8
+size 179754
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..846ad76c14efa4fe4d81cb64f207a4779feb72ec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse of a classic red car parked on a gravel surface. The car is a vintage model with a shiny finish, and it features a distinctive grille and headlights. The license plate reads \"SRS 873K\". The car is parked next to a black car, and the background shows a blurred image of trees and foliage, suggesting that the location is a park or a similar outdoor setting. The style of the video is realistic and it captures the details of the car and its surroundings with precision. The time-lapse effect gives a sense of the passage of time, as the car remains stationary while the background changes subtly."
+    ],
+    "video_ids": [
+        "CYcZKZjVT6U_60_0to112"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A classic red vintage car with a shiny finish, distinctive grille, and headlights. The license plate reads 'SRS 873K'. Next to it, there is a black car.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a classic red vintage car with a shiny finish, distinctive grille, and headlights, and the license plate clearly reads 'SRS 873K'. A black car is visible in the background to the left, matching the description. The video successfully fulfills the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a classic red vintage car with a shiny finish, a grille, and headlights. which the license plate reads reads 'SRS 873K'. There black car is also next the background, the left of fulfilling the description. The setting largely fulfills the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1ec20a1ac37cc67ac5ff4d15742846f13aac9d2b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f8afd088ac80b99284a8b9dd06a050703b5882d2f1e05cda41ea353a9f7d742d
+size 223640
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..c49292aadfffe2591bd70a626aacf8284401c942
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a black and gray computer tower on a wooden desk. A silver coin is placed on the front of the tower, highlighting its size and design. The coin is positioned in the center of the tower, drawing attention to its details. The background is blurred, emphasizing the computer tower and coin. The style of the video is a simple yet effective demonstration of the computer tower's size, using the coin as a reference point. The focus is on the computer tower and coin, with no other objects or actions in the video."
+    ],
+    "video_ids": [
+        "561vc-JUWHQ_6_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black and gray computer tower, a silver coin? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a black and gray computer tower with a silver coin placed next to it. The coin is visible and identifiable as a U.S. quarter, and the computer tower's design, including its ports and ventilation patterns, is consistent with the description. No elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a black and gray computer tower with a silver coin placed in to it. The presence is positioned and positioned as silver silver.S. quarter, which it computer tower is design and including its color and the,, matches consistent with a description. There additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f215c799ff9c8191ba9a191522b0b35c521d5037
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0000/sample-1279-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ce0035800843f07adfec347533ae3f7996b780ff80a9010635af60ec9eaac0b
+size 94778
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..b3574c8f3bea997ffb33f3cb116eecc4d05ff227
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a scene from a car show, featuring a man in a blue shirt and glasses sitting in the passenger seat of a car. The man is engaged in a conversation with the driver, who is not visible in the frame. The car is a sleek, black model with a curved design, and the interior is well-lit, highlighting the man's glasses and the car's interior. The setting appears to be a car showroom or a similar venue, with other cars visible in the background. The style of the video is a mix of documentary and promotional, with a focus on the car and the man's experience in it. The video captures the man's reactions and comments on the car, providing a glimpse into the car's features and performance."
+    ],
+    "video_ids": [
+        "XsIg2cvcaxE_41_0to167"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue shirt and glasses, a sleek, black car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a blue shirt and glasses seated inside a sleek, black car. The setting and attire match the description, and no elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing a blue shirt and glasses, inside a sleek, black car. The man appears the of the description provided and there additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e48367c6fa0f99b386fae4b8f1315581762a0d86
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3da7fa2ec1b9b950db998390ae1599571a195867d011e4bb9ef1528736f22b6c
+size 191754
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..e9de4123523ea9d4e322b39cc5f38fff16593c8c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a single car wheel in a dark room. The wheel is silver with a shiny finish and has a five-spoke design. The tire is black and appears to be new. The wheel is positioned in the center of the frame and is the only object in the video. The style of the video is a simple, straightforward product shot with a focus on the wheel's design and details. The lighting is soft and even, highlighting the wheel's features without creating harsh shadows. The background is dark and nondescript, ensuring that the viewer's attention remains on the wheel. The video does not contain any text or additional objects. The overall impression is that of a clean, professional product presentation."
+    ],
+    "video_ids": [
+        "O_8TviYBU_4_14_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Single car wheel, tire? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a single car wheel with a tire mounted on it, centered in the frame against a dark background. The focus remains entirely on the wheel and tire throughout the video, with no additional conflicting elements present. The description 'Single car wheel, tire' is accurately fulfilled.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a single car wheel with a tire, on it. which in the frame. a background background. The wheel is on on the wheel and tire, the video, fulfilling no additional elements elements present. The description ofSingle car wheel, tire' is fulfilled represented by}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a956c4183ee3860a2cb7d9966b7915a61f991205
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d60bc39e69ed80e348137c26e0050878405ea6d53aa66924a4baafa2a523f339
+size 85110
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..dcf03933f49c80b4c3fc6037b6379e67a2785422
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a well-organized computer desk setup. The desk is white and holds a black computer monitor, a black keyboard, a black mouse, and a pair of black headphones. The monitor displays a colorful graphic of a dragon. To the left of the monitor, there is a microphone on a stand. On the right side of the desk, there is a smartphone in a red case. The desk is situated in a room with a white wall. The style of the video is a simple, straightforward documentation of the desk setup, with no additional action or movement."
+    ],
+    "video_ids": [
+        "rIv9rIL31i8_15_0to164"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: black computer monitor, black keyboard, black mouse, black headphones, colorful graphic of a dragon, microphone on a stand, smartphone in a red case? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black computer monitor displaying a colorful graphic that resembles a dragon or stylized creature, a black keyboard, a black mouse, black headphones, a microphone on a stand, and a smartphone in a red case. All specified objects are present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black computer monitor, a colorful graphic of resembles a dragon. aized creature. a black keyboard with a black mouse, black headphones, a microphone on a stand, and a smartphone in a red case. The the objects are present and match the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6c8fe46475c37a2a1c7c620bd8f1cd83bd7ef2a3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:275ababaa182ed0d22d944759290de8ead332a6ce2b00c77b50b8b53293bf017
+size 58218
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..d8c144d577d45bda6707acfdab901a30c24e5afc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a journey on a winding mountain road. The road, painted with a yellow center line, curves gracefully around the mountain, disappearing into the distance. The road is flanked by a sturdy metal guardrail on the right side, providing a sense of security for the travelers. The surrounding landscape is a mix of lush green trees and towering mountains, their peaks dusted with snow. The sky above is a clear blue, dotted with fluffy white clouds. The overall style of the video is a dynamic and immersive travelogue, capturing the beauty and grandeur of the mountainous landscape."
+    ],
+    "video_ids": [
+        "9I2apS01YlA_123_0to184"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Winding mountain road, yellow center line, metal guardrail, lush green trees, snow-capped mountains, clear blue sky, white clouds? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a winding mountain road with a yellow center line, metal guardrails on both sides, lush green trees lining the road, snow-capped mountains in the background, and a clear blue sky with white clouds. All specified elements are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a winding mountain road with a yellow center line, a guardrails, the sides, lush green trees along the road, snow-capped mountains in the background, and a clear blue sky with white clouds. The these elements are present and contribute depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9cddefd5d05f1484595d3fdc5123c53f2b8a57d5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1279-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:35257082dd792e064a4010b672f654c4e91d6220fdb173f59fe817996bcbf12d
+size 269657
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..fdf3fa9c5042303912dee64e6d041617471a6fa9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene aerial view of a winding road that hugs the shoreline of a deep blue lake. The road, curving gracefully, is flanked by lush green trees and bushes, providing a natural canopy over the asphalt. The lake, a stunning shade of blue, stretches out into the distance, its calm waters reflecting the surrounding landscape. In the distance, majestic mountains rise up, their peaks touching the clear sky. The video is a beautiful representation of a tranquil journey through a picturesque landscape, with the road serving as a guide through the natural beauty of the area."
+    ],
+    "video_ids": [
+        "XA2HBBrkUbU_4_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Winding road, lush green trees and bushes, deep blue lake, calm waters, majestic mountains? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a winding road along the lakeshore, lush green trees and bushes lining the road and surrounding the lake, a deep blue lake with calm waters, and majestic mountains in the background. All core elements described are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as showcasing depicting a winding road that the edgeideore, lush green trees and bushes on the road, the the lake, deep deep blue lake with calm waters, and majestic mountains in the background. The these elements of in present and contribute depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..835c17094b297a72c4e34f8aecdb1ad63e3a83e8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:29238eb865a9cd2f175eaf8008778cae9a61c7f5303b9018de8722a345dea087
+size 249567
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..2f24bf0f54f549d642c1e0a1865c90ee900b4c46
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene at a shallow water's edge, likely a tidal flat or a riverbank. The foreground is dominated by a rocky shore scattered with pebbles and small stones, creating a textured surface. In the middle ground, a bird, possibly a wading bird such as a sandpiper or a similar species, stands in the shallow water. The bird is preening itself, moving its head and neck rhythmically as it grooms its feathers. Its body is partially submerged, with only its legs and part of its back visible above the waterline. The bird's plumage is a mix of white and gray, blending subtly with the muted tones of the surrounding environment.  The background features more rocks and some dried vegetation, suggesting a natural, undisturbed habitat. The water appears calm, with gentle ripples reflecting the overcast sky above. The overall atmosphere is tranquil, with no other animals or significant movement observed in the frame. The camera remains stationary throughout the sequence"
+    ],
+    "video_ids": [
+        "d0323a2f0309889d532954b3f62c39e8c34b624393942e32924090f5272708d1"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bird, possibly a wading bird like a sandpiper.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bird standing in shallow water, which matches the description of a wading bird like a sandpiper. The bird's posture, long legs, and environment are consistent with typical wading bird behavior. There are no elements in the video that contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bird that in shallow water, which is the description of a wading bird like a sandpiper. The bird's posture and with legs, and the are consistent with the wading bird behavior. The are no elements in the video that contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5457a71f454e999fbaae6b887dcd7b0b0e126c2c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e1aabbd311a60d887255dff137fa2fa33c482b48adc48494d87822a7f32a2744
+size 220722
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..9a4764b2876e2c1d2d890f61d213fecf9439efd4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up shot of a plate of food, featuring a green bowl filled with a colorful salad, which includes chunks of avocado, tomato, and onion. The salad is garnished with a lime wedge, which is placed next to the bowl. The plate is set on a wooden table, and there are tortilla chips scattered around the plate. The style of the video is simple and straightforward, focusing on the food and the textures of the ingredients. The lighting is bright and even, highlighting the vibrant colors of the salad and the freshness of the lime. The overall impression is one of a healthy and appetizing meal, ready to be enjoyed."
+    ],
+    "video_ids": [
+        "Z1GCwPmOtMo_17_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Green bowl, colorful salad (avocado, tomato, onion), lime wedge, tortilla chips? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a green bowl filled with a colorful salad containing avocado, tomato, and onion, accompanied by a lime wedge and tortilla chips on a plate. All the specified objects are present and accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a green bowl containing with a colorful salad consisting chunks, tomato, and onion. a by a lime wedge and tortilla chips on a plate. The the elements objects are present and match depicted, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..41a33d2f6f29a5d97df25e9c58c45d5281eab6ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1db404bcbac6c663f42021b3b003a92626f5b0e113e826fd5c5dd2e3834de5ce
+size 81283
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..c80e000ed1b76c7bc8e75096a16854f21464d619
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene day at a marina, where a sleek black speedboat is docked. The boat, with its polished exterior and large windows, is the focal point of the scene. It's moored next to a white dock, which extends into the calm blue water. The marina is bustling with activity, with several other boats visible in the background. The sky above is a clear blue, suggesting a sunny day perfect for boating. The overall style of the video is a simple yet beautiful depiction of a day at the marina, with the speedboat as the star of the show."
+    ],
+    "video_ids": [
+        "QmJY9PtPp5M_45_0to162"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A sleek black speedboat, a white dock, several other boats.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a sleek black speedboat docked at a white pier, with several other boats visible in the background. The scene matches the described elements without contradiction, even though additional elements like a blue mat and a white inflatable boat are present, which do not conflict with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a sleek black speedboat docked at a white dock. which several other boats visible in the background. The setting is the description elements, any.\"\n fulfilling though the elements like the mar sky on a yellow boat boat are present, they do not conflict with the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b5973916f97625fba187363bca481d722c5a6d25
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d379c813f64c0db3a56a266123cf33767c3c47451853111038082b874f552b50
+size 142256
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..3a7ba4ce12fcaaf39d2277f5b92888e09e6b72fe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and a hat, standing in a room with a white wall and a colorful tapestry hanging on it. The man is smiling and appears to be in a good mood. The room has a window with a plant in it, which adds a touch of greenery to the scene. The overall style of the video is casual and relaxed, with the man as the main focus. The room's simple decor and the man's casual attire suggest a laid-back atmosphere. The colorful tapestry adds a pop of color to the otherwise neutral-toned room. The window with the plant provides a nice contrast to the indoor setting. The man's smile and relaxed demeanor suggest that he is enjoying his time in the room."
+    ],
+    "video_ids": [
+        "YjahgcyKXZI_15_190to350"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and a hat, a colorful tapestry, a plant in a window, the man's smile? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man with a beard and a hat, a colorful tapestry hanging on the wall behind him, a plant visible near a window on the left side, and the man smiling at various points. All elements described in the condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man with a beard and a hat, a colorful tapestry on on the wall, him, a plant in in the window, the left side, and the man is. the points. The these in in the condition are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..398111be654f3eda79d583dba035216d7c89ff5f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1dcba7ab724ce34c4cd489cb5a90892489653c6d565f96f997eb52868d34d405
+size 171590
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..e96c76606bd1cead4eaea437c671d05a43f90a93
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a journey on a winding road through a rural landscape. The car's side window is open, revealing a clear blue sky and a lush green field. The road curves gently, and the car's side mirror reflects the surrounding scenery. The car is moving at a moderate speed, and the driver's hand is visible on the steering wheel. The overall style of the video is a realistic depiction of a road trip, with a focus on the natural beauty of the countryside."
+    ],
+    "video_ids": [
+        "DqVsTrXNKCo_31_0to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A car, a clear blue sky, a lush green field, and the car's side mirror.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a car (visible from the side), a clear blue sky with some clouds, a lush green field alongside the road, and the car's side mirror (visible on the passenger side). All core elements of the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a car frompart through the interior mirror a clear blue sky, a clouds, a lush green field, the road, and the car's side mirror reflectingreflect in the left side). The these elements of the description are present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5ad407b23103e1a0189760fdabf5b7932918ec5b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a514c127df5b83ce5308830c296e8770bcb3a3becddb768ef0c3bc5720c8eadd
+size 123192
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..8a847ada3d5a6ddc3d82d608fe846333ec28656e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a young woman standing in a crowd, smiling and looking to her left. She is wearing a black and white patterned top, a green jacket, and a black necklace with a feather pendant. The crowd around her is diverse, with people of various ages and styles of clothing. The setting appears to be outdoors, possibly during a public event or gathering. The woman's expression is cheerful and engaged, suggesting she is enjoying the event or conversation. The style of the video is candid and natural, capturing a moment in the woman's day without any apparent staging or posing."
+    ],
+    "video_ids": [
+        "IrxTtqa24M8_32_0to180"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman standing, smiling, and looking to her left. She is wearing a black and white patterned top, a green jacket, and a black necklace with a feather pendant.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman who is standing, smiling, and looking to her left. She is wearing a black and white patterned top, a green jacket, and a black necklace with a feather pendant. The description matches the visual content accurately.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman standing is standing, smiling, and looking to her left. She is wearing a black and white patterned top, a green jacket, and a black necklace with a feather pendant. The description matches the video elements of.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7535bb53d6aa6f0af3e57e109f181140cfaeb41e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1281d7b286a23451ce620bd9c12bc7d2d6b862f3f94b2ca0e79fd9b687f19914
+size 324035
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..f13c6aef1d51b8ca76a86603c3348e07ae0baa21
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a small, colorful bird perched on a branch in a lush, green forest. The bird, with its vibrant blue head and green body, is the main focus of the video. It is seen in three different positions on the branch, showcasing its agility and grace. The bird's long tail feathers are also visible, adding to its striking appearance. The background is filled with the verdant greenery of the forest, providing a natural and serene setting for the bird. The video is a beautiful representation of nature and wildlife, capturing the bird in its natural habitat."
+    ],
+    "video_ids": [
+        "_hKvS8p3JCM_36_138to292"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small, colorful bird with a vibrant blue head and green body, and long tail feathers.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small, colorful bird with a vibrant blue head and green body, and long tail feathers, which matches the description. The bird is perched on a branch, and its distinctive features are clearly visible throughout the frames.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a small, colorful bird with a vibrant blue head and green body, which long tail feathers, which matches the description provided The bird is perched on a branch, and the colors color are clearly visible, the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b40f4e75f9963b0c674de22b6d1ff9d7b2e460d0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-128-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e423fa169787a48fd99e73347b2b34a92f57168d8ba4578242f057b493d9c180
+size 119007
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..4c4beb9783c407291aeeff038f067c1e8888bef6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a blue car parked on a gravel surface. The car has a sleek design with a black tire and a blue rim. The car is stationary and there are no people or other objects in the frame. The style of the video is a simple, straightforward shot of the car without any additional context or action. The focus is solely on the car and its design."
+    ],
+    "video_ids": [
+        "Uxik0fZO0Fo_26_104to291"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue car with a sleek design, black tire, and blue rim? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a blue car with a sleek design, featuring black tires and blue rims (with blue lug nuts visible). These elements match the description provided. The presence of additional features like a rear spoiler and roof rack does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close car with a sleek design, which a tires and blue rims.al a accents nuts).). The elements match the description provided, The car of gravel elements like the gravelview and the bars does not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..456cac8eb054118966147fa745ddcd0298bef855
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03239d6c40cd6231da03449668d5f90bb7d44def228b7fdb04a1ccfd22cfddd8
+size 183905
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..73bed5887c11300c04355713ca50e5352b10a985
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of triumph and camaraderie among a group of football players. The central figure, a player wearing a red jersey with the number 27, is the focal point of the scene. He is in the midst of a celebratory gesture, his arms outstretched in a victorious pose. His teammates, also dressed in red jerseys, surround him, their faces lit up with smiles as they share in the joy of the moment. The background is a blur of green, suggesting they are on a football field. The overall style of the video is dynamic and energetic, capturing the intensity and excitement of the sport."
+    ],
+    "video_ids": [
+        "ONs2OPfG2GM_67_105to318"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a red jersey with number 27, and his teammates in red jerseys.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a player in a red jersey with the number 27 prominently displayed in the foreground, surrounded by teammates also wearing red jerseys. All players are consistent with the description, and there are no conflicting elements that contradict the core requirement.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a player in a red jersey with the number 27, displayed. the center. and by other also in red jerseys. The players are in with the description, and the are no elements elements.\"\n contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4fb377e6e848139c2a1e653a8fa701e7bf84ea28
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d9d1bb68be34337984ef898ce5db6cd0568f39b955aa6919c009ae3872b53830
+size 218671
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..eb22efadb02f84a2cb776abf03100ab77b4d8be5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of disappointment and reflection for a group of football players. The players, dressed in white and blue uniforms with the number 24 prominently displayed, are seen walking off the field. One player is seen wiping his face with his hand, possibly in frustration or sadness. The other players are seen looking down, their heads bowed in defeat. The background is a blur of green, suggesting a stadium filled with spectators. The overall mood of the video is somber and reflective, capturing the aftermath of a lost game."
+    ],
+    "video_ids": [
+        "Js68KWJkSvw_19_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Football players dressed in white and blue uniforms with the number 24 prominently displayed.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows football players in white and blue uniforms, and the player on the left clearly has the number 24 prominently displayed on his jersey. The description is largely accurate, with no major contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows football players wearing white and blue uniforms with with the number in the left has has the number 24 prominently displayed on his jersey. The players is largely matched, and the significant contradictions.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..18d0a4685076ffd8584d02c15e866fff1b2b3541
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4a680c2509fa7f0ecc42f80665102d714729b13424d31c9d5e818e95c679e611
+size 388000
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..47b39648f8d3a5bdaebc2d3ed95a99860f719d6c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment in a hockey game, featuring a player in a red jersey with the number 8. The player is wearing a red helmet and has a beard. He is standing on the ice, looking off to the side, possibly observing the game or strategizing. The background shows a hockey rink with red seats, indicating that the game is being played in a stadium. The style of the video is a standard sports video, capturing the player in action during the game."
+    ],
+    "video_ids": [
+        "NF41PcRU8tg_15_0to196"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a red jersey with number 8, wearing a red helmet and having a beard.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a hockey player wearing a red jersey with the number 8 visible on the helmet, a red helmet, and has a beard. These elements match the description provided in the 'Object(s)' condition. The presence of other players and background elements does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player player wearing a red jersey with the number 8, on the back. a red helmet, and a a beard. The elements match the description provided, the questionObject(s)' condition.\"\n The player of the elements or the elements does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4de57fd11db2dc9ae49211893174f571da6896a3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90963ba28e4289d6220147411f65ca6f6465b8365b78073bd2cfb3f15ff1830f
+size 190113
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..e80e25556279f50f72873601ee9bf95e3b5d5e32
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen in a workshop setting, showcasing his 3D printing creations. He is holding an orange figurine in his hand, which appears to be a unique design. The workshop is filled with various objects, including shelves stocked with 3D printing materials and a table displaying more of his creations. The man seems to be proud of his work, presenting it with enthusiasm. The video captures the essence of the 3D printing process and the creativity it enables."
+    ],
+    "video_ids": [
+        "BuSQJHZ8P0Q_51_441to609"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, an orange figurine, shelves, 3D printing materials, a table, and more creations.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man holding an orange figurine, with shelves in the background filled with various items, including what appear to be 3D printing materials. There is also a wooden table with multiple 3D-printed creations, including pumpkins and candles, which aligns with the described elements. No core elements are contradicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man holding an orange figurine, which shelves in the background containing with various colorful that including what appear to be 3D printing materials. There are a a table table with more orange3D-printed objects, including akins and other. which align with with the description '. The contradictions elements are missingicted,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c0d2e25bd25bb6185f81e52417a492ec68a8d5d8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7c05436ed54e98f40e21f645438f7c6bf370f39e18642baa9a0ef2af7d7d174d
+size 113710
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..05109be97992e64b04045918ce3fa071c9e63353
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a textured surface, likely a wet pavement or a similar material, with intricate patterns created by water droplets and streaks. The surface is predominantly blue with white highlights, suggesting it might be a painted or treated material. A small insect, possibly a bee, is seen crawling along a thin, dark line that cuts across the surface. The bee's movement is slow and deliberate, indicating it may be navigating through the water or searching for something specific. The camera remains stationary throughout the sequence, focusing on the bee's journey across the textured background. The lighting is natural, casting subtle shadows that enhance the three-dimensional quality of the scene. There are no other significant objects or characters present in the frame, keeping the viewer's attention solely on the bee and its interaction with the environment."
+    ],
+    "video_ids": [
+        "748324cc5df99675165e68ebd5af9d6b80b4404bc4b32678b5f94632a8054a2c"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small insect (possibly a bee) crawling along a thin, dark line.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small, fuzzy, brownish insect that resembles a bee, positioned on a thin, dark line that appears to be a crack or seam in a textured surface. The insect is actively crawling along this line, matching the core description. The surrounding surface has a blueish, textured appearance, but this does not contradict the main subject's behavior or identity.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a close black dark object dark object object-like appears a bee, crawling along a thin, dark line. runs to be a crack in a in a surface surface. The surface is moving moving along this line, which the description description of The surface area is a glossy and, abstract appearance, which this does not contradict the presence focus of description and presence.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f6318ac56872fd98b2128c81d63045e22d7114c0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9cee633ee4a46d6ba7ed2fa0fcaf8000344ace5ae06fa7d88e082a400ca24caa
+size 159934
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..c27ccf082d0c6782894c4ce75f84202fa3427602
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a simple DIY aquarium setup on a wooden table against a plain blue background. The aquarium is a clear glass box placed inside a transparent plastic container labeled \"Craft Master.\" Inside the aquarium, there are several decorative elements: small yellow spherical stones, white pebbles, green aquatic plants, and a small black figurine resembling a penguin. A clear plastic bottle with a blue cap sits atop the aquarium, connected to a white tube that extends into the water, likely serving as an air pump or filter.  A hand enters the frame from the right side, holding a yellow pencil with a white eraser at one end. The hand carefully places the pencil into the bottle's cap, positioning it so that the eraser rests inside the bottle while the pencil extends upwards. The hand then adjusts the pencil slightly, ensuring it is securely placed. The video captures this action in a steady shot, focusing on the interaction between the hand, the pencil, and the bottle. The background remains static throughout"
+    ],
+    "video_ids": [
+        "3fa8afea759cc9bb034e2740a2f366156723665d57d1a15d36d694f8da27a281"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: DIY aquarium, wooden table, plain blue background, small yellow spherical stones, white pebbles, green aquatic plants, small black figurine, clear plastic bottle with a blue cap, white tube, yellow pencil with a white eraser.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by showing a DIY aquarium setup on a wooden table against a plain blue background. It includes small yellow spherical stones, white pebbles, green aquatic plants, a small black figurine, a clear plastic bottle with a blue cap, a white tube, and a yellow pencil with a white eraser. All specified elements are present and correctly identified in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting a DIY aquarium setup. a wooden table with a plain blue background. The includes small yellow spherical stones, white pebbles, green aquatic plants, a small black figurine, a clear plastic bottle with a blue cap, and white tube, and a yellow pencil with a white eraser. The these objects are present and match described in the scene.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..df4e7dd28ff651d537c2ab620373e6c4236d88e8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6b5095c33147dec7b222f4fbc99b4221de42d23cde118925a302625aeb4f1b5
+size 101028
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..25f3d1fc481cca3b85a352c39eb470c7e984b120
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a bowl of pasta, which appears to be a creamy pesto pasta dish. The pasta is served in a white bowl, and the dish is garnished with a sprig of basil. The pasta is creamy and green, suggesting the presence of pesto sauce. The dish is presented in a way that highlights the creamy texture of the pasta and the vibrant green color of the pesto sauce. The basil garnish adds a touch of freshness to the dish. The overall style of the video is simple and straightforward, focusing on the pasta dish without any additional elements or distractions. The video does not contain any text or other objects. The focus is solely on the pasta dish, and the viewer is left to imagine the taste and texture of the dish. The video does not provide any information about the location or setting of the dish, leaving the viewer to focus solely on the pasta dish itself. The video does not contain any action or movement, and the pasta dish remains stationary throughout the video. The video does not contain any sound or music, allowing the viewer to focus solely on the visual presentation of the pasta dish. The video does not contain any text or other objects, and the viewer is left to focus solely on the pasta dish. The video does not contain any action or movement, and the pasta dish remains stationary throughout the video. The video does not contain any sound or music, allowing the viewer to focus solely on the visual presentation of the"
+    ],
+    "video_ids": [
+        "KTadTKRctMw_36_0to112"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: bowl of pasta, white bowl, sprig of basil? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bowl of pasta, specifically fusilli, coated in green pesto sauce, served in a white bowl and garnished with a sprig of fresh basil. All elements mentioned in the description are clearly visible and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close of pasta with which ailli, which in a sauceo.. which in a white bowl. garnished with a sprig of basil basil. The the of in the condition are present present and match represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9db4c51113935aaae55b3df66c328d714b0e3103
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1280-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb56636c7003d6659d08fd61f0f9056334ae67300df607c0e6f46f0843de377f
+size 134289
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..10473f40dae8f93a9e89f0b6c4f701bc666686d4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video begins with a close-up shot of a pink balloon adorned with a cat face design, resting on a snowy surface. A hand holding a knife approaches the balloon from the right side of the frame. The knife is positioned near the balloon's opening, and as it makes contact, the balloon bursts, releasing a vibrant red liquid into the snow. The liquid spreads out, creating a vivid splash against the white backdrop of the snow. The camera remains stationary throughout the sequence, focusing on capturing the moment of the balloon's burst and the resulting colorful explosion. The scene is set outdoors, likely during winter, given the presence of snow. There are no other objects or characters visible in the frame, keeping the focus solely on the balloon and the knife."
+    ],
+    "video_ids": [
+        "ed563ceff02846d208d7f31969ee45920d8cdb1748e6aec2abe86081756cf57b"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Pink balloon with a cat face design, knife, vibrant red liquid, snow? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a pink balloon with a cat face design, a knife being used to puncture it, and the subsequent release of a vibrant red liquid onto snow. All core elements described are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful the pink balloon with a cat face design, a knife, used to cuture the, and the resulting release of vibrant vibrant red liquid. a. The these elements of in present, match represented in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7bf6975595ebe0643957e35beb87fbf21f99a360
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0d6887730e59370a3d48a0c84397668ac486a35991d997964fcf73d01695e533
+size 123739
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..95193266ce361576abf7a35633b819ae784c90fe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a cup of instant noodles being prepared. The cup is placed on a table, and water is being poured into it from a kettle. The water is clear and appears to be hot. The cup is white with a colorful label that reads \"Salted Egg\". The label is bright and colorful, with a yellow background and red and green text. The kettle is silver and has a black handle. The water is being poured from the kettle into the cup, and the cup is filling up with the hot water. The video captures the process of preparing the instant noodles, from the pouring of the water to the filling of the cup. The focus is on the cup and the kettle, with the background being out of focus. The video is a simple and straightforward depiction of a common activity, with no additional elements or distractions."
+    ],
+    "video_ids": [
+        "vQvMbUhAeGM_55_108to231"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A cup, a kettle, and a table.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a cup of instant noodles, a kettle pouring water into it, and a table on which the cup is placed. These objects are present and consistent with the described condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a cup being instant noodles being a kettle pouring hot into the, and a table with which the cup and placed. The elements fulfill present and fulfill with the description '.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..09fbae9e2b940c5148edf1a3eae9f8db5f120da4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2df8566054954ecc0b6398c3b0959b67344c3a1b156975b44dc7fa0354414aa1
+size 114496
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..916b93c185f915e42a74df4a3fb0dd173a434b55
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two women standing in a snowy environment, dressed warmly for the cold weather. They are wearing winter coats, hats, and scarves, and appear to be posing for the camera. The women are smiling and seem to be enjoying themselves despite the cold. The background is a snowy landscape with trees and a sign that reads \"Blumen Kr\u00fcschen\". The style of the video is casual and friendly, capturing a moment of joy and camaraderie between the two women in a winter setting."
+    ],
+    "video_ids": [
+        "p4u95GteAjQ_5_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two women dressed in winter coats, hats, and scarves, smiling and posing.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two women dressed in winter coats, hats, and scarves, and they are smiling and posing for the camera. The background elements, such as the sign and snow, do not contradict the core description and are consistent with a winter setting.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two women dressed in winter attire, hats, and scarves, and they are smiling and posing. the camera. The description appears, such as the snowy and the, do not contradict the description description and are acceptable with a winter setting.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cf46c61d08c394898362c4507b9d7d0404c15592
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e421afbd491dc0ebe8ab47b59031eae568dbed52c59d312e8e1fb5876a21a662
+size 258101
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..97471cefe6125a3bfedc5a7fa517ec5b58038f3e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person using a yellow and black power saw to cut through a piece of metal. The saw is held firmly in the person's hand, and the blade is spinning rapidly. The metal being cut is placed on a sturdy workbench, and the person appears to be concentrating on their task. The background of the video shows a parked car and a leaf-covered lawn, suggesting that the scene is taking place outdoors. The style of the video is straightforward and documentary, focusing on the action of the saw cutting through the metal."
+    ],
+    "video_ids": [
+        "-FP3MmF0GcI_16_0to190"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a yellow and black power saw, and a piece of metal on a workbench.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person interacting with a yellow and black DeWalt power saw, which is mounted on a workbench. Although the saw is not actively cutting metal, the person is handling the saw blade and the setup suggests metal cutting is intended or in progress. The presence of a metal workbench surface is also evident. The description does not require the metal to be actively cut, only to be present on the workbench, which is satisfied.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person using with a yellow and black powerWalt power saw, which is positioned on a piecebench. The the video is not cutting cutting a in the presence's holding the saw,, appears work suggests theworking is about. has progress. The presence of a work workbench and and also visible, The video of not specify the saw to be actively being, only that be on on the workbench, which is the by}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a90f17c31fffb6d39629aab599d15656775c8325
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:65f0ac28fd2dc00a0bd6a700eca452e13009b7175198e16e691b3ad692c50e65
+size 176293
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..28c775a8cbdfdc1247cfca1723fa80f1bc9420b3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman standing in front of a wooden wall. She is wearing a white hijab and a black top. In the first frame, she is holding a small object in her hands. In the second frame, she is still holding the object, but her hands are now clasped together. In the third frame, she is no longer holding the object, and her hands are at her sides. The style of the video is simple and straightforward, focusing on the woman and her actions."
+    ],
+    "video_ids": [
+        "CzHP-PIFBH0_3_0to181"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a wooden wall, a small object? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman wearing a hijab, standing in front of a wooden wall. She is holding a small object, which appears to be a remote or clicker. All elements described in the 'Object(s)' condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a woman standing a hijab, standing in front of a wooden wall. The is holding and small object in which appears to be a piece or aer, The the of in the 'Object(s)' condition are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6b81e32c493515d074fb6e766a4c42c0d785b27d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:940956414b04804a16cec71bd377b74293c45abf8cfced8494591ee8a6a63cd3
+size 110937
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..9059aa3f68568e2b85ae629019dca9186138f20d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen interacting with a large tortoise on a tiled floor. The tortoise is covered in a white substance, possibly dust or sand, and appears to be in motion. The person is gently touching the tortoise's shell, possibly in a soothing or comforting manner. The tortoise seems to be responding to the touch, as it is opening its mouth and making a sound. The tiled floor provides a contrasting texture to the smooth shell of the tortoise. The interaction between the person and the tortoise creates a sense of connection and care. The video captures a unique moment of human-animal interaction, highlighting the gentle nature of the person and the curious behavior of the tortoise."
+    ],
+    "video_ids": [
+        "6LBzeOs-nNE_19_317to463"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a large tortoise? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person interacting with a large tortoise on a tiled floor. The tortoise is being gently rubbed by the person, and its features are consistent with a large tortoise. The presence of a paint can and a smartphone does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a person interacting with a large tortoise. a pattern floor. The personoise is the held handled under the person's which the shell, visible with a large tortoise. The presence of the tiledbrush and a paint in not contradict the core description of}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cd6db0a2b04f4ebb3b0eb59c3a03b439349fbda9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:824d48f0196d2977db593faef20615fd07b9d6f65c80467100a815c094b483a0
+size 219805
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..63bf41e009cb27f85a96e79d985d2d1aedb4ab6c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video opens with a serene view of a lush, verdant forest, where dense green foliage dominates the landscape. A waterfall cascades down a rocky cliff in the background, adding a dynamic element to the tranquil setting. The camera remains stationary throughout the sequence, allowing viewers to take in the natural beauty of the scene. In the foreground, a river flows gently, reflecting the surrounding trees and sky. A bird, possibly a heron or egret, is seen flying gracefully over the water, its wings spread wide against the backdrop of the forest and waterfall. The bird's movement is smooth and fluid, creating a sense of harmony between the animal and its environment. The overall atmosphere is peaceful and untouched, capturing the essence of nature in its raw form."
+    ],
+    "video_ids": [
+        "8ae68cbf5c4662cd2d5d3898ab00efc6a52e1d0e0782fe8ee6587f1f2a33fe45"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bird, possibly a heron or egret, flying gracefully over the water.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bird, consistent with the appearance of a heron or egret, flying gracefully over a body of water. The bird's flight path and the surrounding lush greenery and water align with the description. The presence of trees and a fallen log does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts a bird, which with the description of a heron or egret, flying gracefully over the body of water. The bird is movement path and the context environment greenery and waterfallfalls with the description of The presence of the and a waterfall log in not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..83ce0bd57d523b3be9d94f8c54e5e9d0bddc03a9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:342aa5c5b35ddc84948bb2d17f876b0f0d5a1edc0cea4446c59925f7d0ca732a
+size 106399
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..e70f66c261ec323432cc6551e7b49ea0108a8aa8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man named Steve, who is a news anchor. He is dressed in a suit and tie, and appears to be in a studio setting. The background is blue, and there is a logo for the news network. Steve is looking off to the side, and seems to be in the middle of a broadcast. The style of the video is professional and polished, typical of a news program."
+    ],
+    "video_ids": [
+        "5U7n-HGabOE_1_0to142"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man named Steve, dressed in a suit and tie, looking off to the side.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man who appears to be Steve Harvey, dressed in a suit and tie, and he is looking off to the side. The background and on-screen text confirm his identity as Steve, and his attire matches the description. The presence of the show's logo and a lamp in the background does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed appears to be dressed,, dressed in a suit and tie, with he is indeed off to the side. The description suggests attire-screen graphics suggest that identity as Steve Harvey and the attire and the description. The video of the on's logo in the blurred in the background does not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bdfac1cb64c8aa28d44a6593a5609d225554ae6d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1281-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:db17f62fe16cd11ca88f464d6fb19490beaa0d6bdfb4255a188980cd72cba078
+size 166337
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..98a88c0bf69bd3b8766e728bb41fb0afb9b8b82f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a young man named Baptiste Noel from the ski club La Bressaude. He is standing in front of a crowd of people, possibly at a ski resort or a similar outdoor event. The crowd is diverse, with individuals of various ages and attire, suggesting a public gathering or event. The background shows a picturesque mountainous landscape with houses and buildings, indicating that the location is likely in a mountainous region. The style of the video is a standard news report, with a focus on the interviewee and the event taking place. The video likely includes an introduction, the interview, and possibly some footage of the ski club or the event."
+    ],
+    "video_ids": [
+        "a7PFj683kNs_14_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man named Baptiste Noel, a crowd of diverse individuals of various ages and attire.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a young man identified as Baptiste No\u00ebl from Ski-club La Bressaude, and there is a visible crowd of diverse individuals of various ages and attire in the background. The scene matches the described conditions without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a young man in as Baptiste Noel\u00ebl, the AssClub Assusresseane,, as there is a crowd crowd of diverse individuals of various ages and attire in the background. The presence is the description elements without any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2c7fc42b4ee2584a2b711e434f6dd920e436ddda
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17dcb3169054b73f016cb0b3b287540ffcce444591a5e26de6d887373d55c36a
+size 107009
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..6d80c449dd715efa84e1ec82085454213484def3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a sports talk show featuring three men in a studio setting. The host, wearing a purple shirt, is seated at a desk with a microphone, engaging in conversation with two other men. The first guest, wearing a suit, is seated to the host's right, while the second guest, wearing a black shirt, is seated to the host's left. The studio has a large screen in the background displaying the Super Bowl logo and the text \"LIVE ATLANTA\". The show is titled \"The Stephen A. Smith Show\" and the host's Twitter handle is displayed at the bottom of the screen. The overall style of the video is professional and polished, typical of a sports talk show."
+    ],
+    "video_ids": [
+        "5AccDDowo6k_10_620to759"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Host (wearing a purple shirt), First Guest (wearing a suit), Second Guest (wearing a black shirt)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three individuals at a sports talk show set. The host on the right is wearing a purple shirt, the first guest in the middle is wearing a suit, and the second guest on the left is wearing a black shirt. All three individuals match the described attire, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows two individuals seated a desk news show set. The person is the left is wearing a suit shirt, and first guest in the middle is wearing a suit, and the second guest on the left is wearing a black shirt. The the individuals match the description attire, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..06f1d61b25b7d33cff9c78d6c55c167606520785
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c56bf7583264307823683a83e9bbe8c8ee6411342d9ef6e919f1d1b00e69fe0e
+size 79214
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..bfdd048b8c1852342b65ef40e64551cdade15634
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a lively scene at a Chiefs football game. The first frame shows a group of fans, all dressed in red jerseys, walking towards the stadium. They are carrying drinks and snacks, ready for a day of cheering. The second frame shows the same group of fans, now inside the stadium, standing in line at a concession stand. The third frame shows the fans, now seated in the stands, watching the game with excitement. The video is a snapshot of the camaraderie and enthusiasm that comes with being a Chiefs fan."
+    ],
+    "video_ids": [
+        "VoyHAzF5xE0_8_0to164"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Fans, drinks, snacks, concession stand, stadium stands? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fans wearing Chiefs jerseys, a concession stand with drinks and snacks, and the stadium stands in the background. These elements align with the described conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a in red jerseys, which concession stand with drinks and snacks, and a interior stands. the background. These elements align with the ' ',}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7ffd59cfc2496bb45af3bfb2d0a39eb247665e18
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e8c751f81c2085590ecd77f8b95210d9e4ebf5181a6af0c339ff1bd62887df7
+size 306156
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..e4bac04cafdf6bf6d74cca35d6427953f7c806b7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene journey through a narrow, mossy cave. The viewer's perspective is from the water, moving down the center of the cave. The cave's walls are covered in lush green moss, creating a vibrant contrast with the calm, light blue water. The cave's ceiling is low, with occasional stalactites hanging down, adding to the sense of depth and mystery. The journey continues, the water's surface undisturbed, reflecting the cave's interior. The video is a testament to the beauty of nature's hidden gems."
+    ],
+    "video_ids": [
+        "_EpeSyqnsfo_216_0to150"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Moss, water, stalactites? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows moss covering the rocks, flowing water in a canyon-like setting, and stalactite-like formations hanging from the ceiling of the canyon. All three elements mentioned in the condition are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a-covered the cave, water water, the cave, setting, and stalactites-like formations hanging from the ceiling. the cave. These the elements are in the ' are present and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..21cbf2d9ee3ced5527645357e1f0f259b1e140de
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a64e68dc5b2983cf3098bab8817548331dd897ab737cd1af8e96d423490775c5
+size 211495
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..701ab5329f65cefde42bbefd07011e7d4e7abd12
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a sleek, green sports car with a number 12 on its side, driving on a gray road. The car is captured from a side angle, emphasizing its aerodynamic design and the number 12 on its side. The car's black rims and the number 12 are clearly visible. The car is in motion, suggesting speed and performance. The gray road contrasts with the green of the car, making the car the focal point of the video. The video is likely a promotional or advertisement video for the sports car, highlighting its design and performance."
+    ],
+    "video_ids": [
+        "L95AjaEAoJI_25_26to177"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A sleek, green sports car with a number 12 on its side.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a sleek, green sports car with the number 12 prominently displayed on its side, matching the core description. Additional elements like the large rear wing and racing tires are consistent with a race car and do not contradict the primary description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a sleek, green sports car with a number 12 prominently displayed on its side. which the description description provided The elements like the road wheels wing and the decals are also with a high car design do not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3e77d8f32af33341b7221a35e10c38f651847c99
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c17e29f7bbb3b986b74bb9644a05a2ad90527f355ffab6c6b5f80c9340b197b5
+size 177596
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..143dde20ea65d8f419e6dc3aef7a2a772a21bfd9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a white and gray hamster in a colorful, cozy-looking bed. The hamster is seen peeking out from the bed, which is covered with a red and yellow blanket. The bed is placed on a surface covered with hay and small black objects, possibly seeds or pellets. The hamster appears to be exploring its surroundings, possibly looking for food or simply enjoying its environment. The overall style of the video is cute and endearing, capturing the small creature in a comfortable and safe setting."
+    ],
+    "video_ids": [
+        "wIFkbQK8oms_21_289to481"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white and gray hamster? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small rodent with white and gray fur, consistent with a hamster, peeking out from a colorful blanket. The animal's features, including its small size, whiskers, and dark eyes, match those of a hamster. There is no contradiction with the description of a white and gray hamster.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small animalent with white and gray fur, which with the hamster. insideing out from a colorful,. The presence's size, such its size size and furers, and the eyes, align the of a hamster. The are no significant with the description of the white and gray hamster.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..09b51315e0f35bcd88707dcf696c188945477f21
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cdd9ea724335b4979ec316c6c6a114883ce5f50503cb58cc5f75839be99e70ee
+size 260072
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..c8f131937571b60e2e75a73d3c95938caf124a41
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman stands in a kitchen, her hands clasped together as she speaks. She is dressed in a striped shirt and her dark hair falls over her shoulders. The kitchen is well-equipped with a sink, a stove, and a refrigerator. The sink is filled with dishes, and a cup sits on the counter. The stove is white, and a knife block is placed next to it. The refrigerator is large and white, matching the overall color scheme of the kitchen. The woman appears to be in the middle of a conversation, her expression serious as she looks off to the side. The video captures a moment of everyday life, with the woman in the kitchen, surrounded by the familiar objects of her home."
+    ],
+    "video_ids": [
+        "WdgU1DQc2oI_1_405to569"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Woman, sink, dishes, cup, stove, knife block, refrigerator? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman standing in a kitchen. Key objects such as a sink, a cup (on the counter), a stove, a knife block, and a refrigerator are all visible in the background, matching the specified 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a woman standing in a kitchen, There elements such as a sink, a cup,which the counter), a refrigerator ( a knife block, and a refrigerator are visible visible. the background. fulfilling the description objectsObject(s)' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..160dc4376faa8511652a355589c05f71f1deb537
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57391cd310d8f0d2d4bb633e1274072bc2c992f347835bcb1244b9e20f98a803
+size 148106
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..6f02c9c518a439ea6a3822b2838fe73041a4d4d1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen preparing a meal in a modern kitchen. She is wearing a white lace top and a gold necklace, and she has blonde hair. The kitchen is equipped with a stainless steel stove and a sink. On the counter, there are various ingredients such as a bottle of olive oil, a jar of herbs, and a bunch of lettuce. The woman is using a cutting board to chop the lettuce. The kitchen has a clean and bright aesthetic, with white tiles on the walls and a blue backsplash. The woman appears to be enjoying herself as she prepares the meal."
+    ],
+    "video_ids": [
+        "o2cVa5FBX9A_31_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Woman, white lace top, gold necklace, blonde hair, cutting board, bottle of olive oil, jar of herbs, bunch of lettuce? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a white lace top and a gold necklace with blonde hair, standing in a kitchen. She is positioned near a cutting board, with a bottle of olive oil and a jar of herbs visible on the counter. A bunch of lettuce is also present, either on the cutting board or nearby. All elements described in the condition are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a white lace top and a gold necklace, blonde hair. cutting in a kitchen. She is using at a cutting board and and a bottle of olive oil and a jar of herbs visible on the counter. The bunch of lettuce is also present on and on the cutting board or on. The these in in the condition are present and match with the video content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c5634f1d5034538da341711751380410a2d0062c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1282-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0762f3f7dfc62f3fadab74fb1e353e9af93387235b8101f019d75884d72f3056
+size 172667
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..0c708e3be8c3efaed3fab70509f8b15c209eeb60
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a scene of two dinosaurs in a forested area. The first dinosaur is a large, brown creature with a long neck and a row of sharp teeth. It is walking through the trees, its body partially hidden by the foliage. The second dinosaur is a smaller, brown creature with a short neck and a row of sharp teeth. It is walking through the grass, its body partially hidden by the tall grass. The scene is set in a forested area with trees and bushes. The dinosaurs are the main focus of the video, and their movements and interactions with their environment are the main elements of the video. The style of the video is realistic, with the dinosaurs appearing as if they are real creatures in a real environment. The video does not contain any text or other elements."
+    ],
+    "video_ids": [
+        "5WUfl4g0xE0_69_0to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two dinosaurs, one large and brown with a long neck, the other smaller and brown with a short neck.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two dinosaur statues in a lush green environment. The larger one in the background has a long neck and is brown, while the smaller one in the foreground has a shorter neck and is also brown, matching the description. The presence of additional foliage and trees does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts two dinosaurs-like that a forest, environment. One larger dinosaur has the background has a long neck and a brown, while the smaller one in the foreground has a short neck and is also brown. matching the description provided The presence of additional elements and the does not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..41dff60d99a5066f398732d9e83866bd15d6726c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75bef96d253ed37421c6dbf760c04cebd35291ed59715691838550276c52800d
+size 334654
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..bebb37ef504805e7971adc1837025c3990094008
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a mother hen and her chicks in a natural setting. The mother hen, with her black and brown feathers, is seen in the center of the frame, tending to her chicks. The chicks, small and fluffy, are seen huddled around her, seeking warmth and protection. The setting appears to be a dirt ground, possibly a farm or a rural area, with trees and foliage in the background. The mother hen's protective behavior and the chicks' dependence on her are the main focus of the video. The style of the video is a simple, unfiltered capture of a moment in nature, with no additional embellishments or effects."
+    ],
+    "video_ids": [
+        "AWTcXCrWUxQ_3_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Mother hen (black and brown feathers), chicks (small and fluffy)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a mother hen with black and brown feathers sitting on the ground, surrounded by small, fluffy chicks. The core description is accurately represented, with no significant contradictions. Additional elements like dirt, rocks, and foliage do not conflict with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a mother hen with black and brown feathers, on a ground, surrounded by a, fluffy chicks. The mother elements of accurately represented in and no additional contradictions or The elements like the and leaves, and green in not conflict with the main.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5750882a1b0e8835a868804de96ef804d8e6d9a5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:452d1d5faa78389a5db27ec3843689303bc1764cfb288ebcad5f87187c5c96f2
+size 154478
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..c7ef3b83c450823a248884287c1d7c2994c61977
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is a quarterback, number 18, who is in the process of throwing a pass. He is wearing a white jersey with blue and orange accents, and his arm is extended, indicating the action of the throw. The quarterback is surrounded by his teammates, who are also dressed in white jerseys with blue and orange accents. They are in various positions, some standing and others in motion, suggesting a high level of activity on the field. The background is filled with the crowd, indicating that the game is taking place in a stadium. The crowd appears to be engaged in the game, adding to the overall atmosphere of excitement and anticipation. The style of the video is realistic, capturing the intensity and action of the game in a way that is both engaging and immersive."
+    ],
+    "video_ids": [
+        "LTkt3hLfN6o_39_150to300"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Quarterback (number 18), teammates, crowd? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing the quarterback (number 18) in a dynamic pose, surrounded by his teammates in formation, and with a blurred crowd in the background. All specified elements are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a quarterback (number 18) in the white throwing, holding by teammates teammates in white, and with a crowd crowd in the background. The elements elements are present and match with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6937c7a7be1297d6ee4d57680cceee1cface8afe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fc270f2b6e4e667e94e2a843929a87e20837f2bfe5b683f6e386849541edd44d
+size 312819
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..6db9eaf14ce2383428e6f95ce531c269204f9fb5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a golden retriever dog sitting on a bed. The dog is wearing a red and white checkered bandana around its neck. The dog is looking directly at the camera with its mouth slightly open, giving the impression that it is panting or smiling. The bed has white sheets and a gray pillow with white stripes. The background is a plain white wall, which provides a neutral backdrop for the scene. The style of the video is simple and straightforward, focusing on the dog and its immediate surroundings. The dog's expression and the bandana suggest that it might be a pet-friendly environment or that the dog is being dressed up for a special occasion. The overall mood of the video is calm and relaxed."
+    ],
+    "video_ids": [
+        "UmVpw2TisOA_4_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A golden retriever dog sitting on the bed, wearing a red and white checkered bandana, looking directly at the camera with its mouth slightly open.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a golden retriever sitting on a bed, wearing a red and white checkered bandana, and looking directly at the camera with its mouth slightly open. The dog's posture and expression match the description, and no elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a golden retriever dog on a bed, wearing a red and white checkered bandana, and looking directly at the camera with its mouth slightly open. The description's posture, the match the description provided and there additional contradict the given description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..66b96857633f9ed7047c716ab71d977d0bfa3283
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dfd3845a561d8faff4e6c0fee46b69e0acd9ece677a6bd887be40ea48f4c14d4
+size 149410
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..9e0c4504b2869b7d2dbf971e95c72e0950197c95
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a small, light brown mouse perched on a thin branch. The mouse, with its tiny black eyes and whiskers, appears to be cautiously exploring its surroundings. The branch, which is the main focus of the video, is thin and delicate, providing a stark contrast to the mouse's small size. The background is blurred, drawing attention to the mouse and the branch. The overall style of the video is a close-up, slow-motion shot, which allows for a detailed view of the mouse's actions and the texture of the branch. The video does not contain any text or additional objects. The mouse's movements and the branch's position remain constant throughout the video. The video does not depict any actions or interactions between the mouse and the branch. The mouse is simply perched on the branch, observing its surroundings. The video does not contain any sound or music. The focus is solely on the mouse and the branch, with no additional elements or distractions. The video does not contain any text or additional objects. The mouse's movements and the branch's position remain constant throughout the video. The video does not depict any actions or interactions between the mouse and the branch. The mouse is simply perched on the branch, observing its surroundings. The video does not contain any sound or music. The focus is solely on the mouse and the branch, with no additional elements or distractions."
+    ],
+    "video_ids": [
+        "hMQRGtvFuqg_28_94to265"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small, light brown mouse with tiny black eyes and whiskers.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a small, light brown mouse with tiny black eyes and whiskers, which matches the description. Although the video includes additional elements (a split-screen with a scientist and lab work), these do not contradict the core description of the mouse.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a small, light brown mouse with tiny black eyes and whiskers. which align the description provided The the mouse is additional elements likesuch branch screen effect a pink and a equipment), the do not contradict the core description of the mouse.}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2328580ae5bb4cb35c7799efc7893e6c5feb2f3d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4956e67664a21b440268d69f562ce1078f4d4faef7d22298b0b3a3aa524fbbc1
+size 219764
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..8e1b857a59c61bf680e2afb2b8b523647323a1a2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a small, blue robotic insect-like drone flying indoors on a checkered floor. The drone, equipped with two large, transparent wings and a body that glows with a blue light, moves swiftly across the floor. As it flies, the camera follows its movement closely, maintaining focus on the drone while the background blurs due to the speed. The drone navigates around a pillar, showcasing its agility and the smoothness of its flight. The environment appears to be an indoor space with a clean, modern design, featuring white walls and a tiled floor with alternating dark and light squares. The lighting is bright, enhancing the visibility of the drone's features and movements. The drone continues to fly, demonstrating its capabilities in a controlled indoor setting."
+    ],
+    "video_ids": [
+        "e48ab57798fe320c5cbbbf30bf32b2ca83fea8f81b22df88e9ee7922842a69f6"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small, blue robotic insect-like drone with two large, transparent wings and a glowing blue body.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small, blue robotic insect-like drone with two large, transparent wings and a glowing blue body, which matches the description. The drone is seen flying in an indoor environment, and its appearance and movement align with the specified characteristics.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a small, blue robotic object-like drone with two large, transparent wings and a glowing blue body. which matches the description provided The drone is moving in in a indoor environment, and its design and features align with the characteristics characteristics.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..10ae251a188c4e8ba05486133a85878fc60be9a1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ceea8ba5945444ac178362d27bf3a31c6e39d88e510a5527c315099cda05eb92
+size 217235
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..ece414c4e9a8ceda86a70d1ee2a7a73bb4ef5699
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a young girl are standing in front of a large drum set on a stage. The man is dressed in a suit and tie, and he is holding a red stick in his hand, gesturing towards the drum set. The girl is standing next to him, looking at the drum set with interest. The stage is set with a blue curtain in the background, and there are other musicians playing instruments in the background. The video captures the anticipation and excitement of the performance, as the man and the girl prepare to play the drums."
+    ],
+    "video_ids": [
+        "Eaje65Z_Dlk_9_91to228"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, a young girl, a large drum set? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a suit and tie standing next to a large drum set, and a young girl is also present nearby. These core elements match the description. While there are additional elements such as a band in the background and stage lighting, they do not contradict the specified objects and are consistent with a performance or show setting.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a man in a suit and tie, next to a young drum set. with a young girl sitting also present,. The elements elements match the description provided The there are additional elements like as a person member the background, a lighting, they do not contradict the main objects and thus acceptable with a performance setting event setting.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..068e27a83dea90112881f9e7fa4396a2a581d6f4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cde1a0ce95b62f82a98b7e8ef6410e5a4cc3903a9bd277e371d1c262558cface
+size 252166
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..58aebf00eb889a8131fbc543a9659cf7038d04d8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with long hair and a beard, wearing glasses and a suit. He is seated in a room with a blurred background, suggesting an indoor setting. The man appears to be speaking, as indicated by his open mouth and expressive facial features. The lighting in the room is soft and even, highlighting the man's features without creating harsh shadows. The style of the video is a close-up portrait, focusing on the man's face and upper body. The overall mood of the video is calm and professional."
+    ],
+    "video_ids": [
+        "IJ-_ypOOMvY_61_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with long hair, beard, glasses, and a suit.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with long hair, a beard, glasses, and wearing a suit (black blazer over a collared shirt), which matches the description. There are no elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man with long hair, a beard, glasses, and a a suit.black jacketazer and a whiteared shirt and which align the description provided The are no additional in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3268a22380d88e0a458288dd77f7ce3e08a53cf0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1283-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b5c69bb5ff9976e15848ebd0a069b67e1b9cc7b399c03ea6d7ff62a89faaaac
+size 211212
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..ed0198f5688e83018f1941ddac6e34328db3a5e0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a sports competition show featuring two teams, Hammerheads and Sentinels. The teams are standing in a large, industrial-style room with a concrete floor and walls. The Hammerheads team is wearing green shirts, while the Sentinels team is wearing purple shirts. The teams are gathered around a white table, which is likely the competition area. The host of the show is standing in the center of the room, holding a microphone and speaking to the teams. The show has a dynamic and energetic atmosphere, with the teams focused on the competition."
+    ],
+    "video_ids": [
+        "QeCoWs2H_Kc_44_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two teams (Hammerheads and Sentinels), a white table, and a host with a microphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows two teams, Hammerheads (in green) and Sentinels (in purple), a white table in the center, and a host (in black shirt) holding a microphone. The host is interacting with the teams, and the scoreboard graphic confirms the team names. Additional elements like a camera and background decor do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful two teams, theheads andin green uniforms and Sentinels (in purple), standing white table, the center, and a host within purple and) holding a microphone. The presence appears positioned with the teams, which the setting is in the presence names. The elements like the door and a details are not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a49caf6b4718f2521c30c44abd2fc0ddfb3879ea
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8e2f31995b556496133388f279089e009f43b7f6f54b924a4064f302cb6047f9
+size 132824
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..0bf107f96ce4c3a139cbdd6dfc64d4db9ce4b838
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man in a black shirt and sunglasses, standing on a city street, taking a selfie. He is gesturing with his hands, possibly explaining something or expressing excitement. In the background, there are parked cars, including a white and red sports car, and a red brick building. The scene suggests a casual, urban setting, with the man as the main subject. The style of the video is candid and informal, capturing a moment in the man's day."
+    ],
+    "video_ids": [
+        "77KmvkKl5YU_16_0to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: man in a black shirt and sunglasses, parked cars, white and red sports car, red brick building? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a black shirt and sunglasses, parked cars in the background, a white and red sports car (Mini Cooper), and a red brick building. All elements described in the condition are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a man wearing a black shirt and sunglasses, standing cars, the background, and white sports red sports car,which Cooper), and a red brick building. The elements match in the condition are present in match depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5dc256dd863648c81fd9ddcbf1fe55fbabc7b082
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b6dc0058630b12399311ee306cc3607a6cc9d661136e070b154cf705e4da943
+size 210253
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..0eb7966e9be570745e10910cda278eb9073d5f6d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the delightful process of dipping a long, thin, crispy churro into a cup of frothy, creamy hot chocolate. The churro is seen in three stages: first, it is held above the cup, poised for the dip; then, it is partially submerged into the hot chocolate, creating a delightful contrast between the golden-brown of the churro and the rich, dark color of the hot chocolate; finally, the churro is fully immersed, its surface glistening with the hot chocolate, ready to be enjoyed. The cup of hot chocolate is placed on a wooden table, adding a rustic charm to the scene. The entire process is captured in a close-up shot, focusing on the churro and the hot chocolate, making the viewer feel as if they are part of the experience."
+    ],
+    "video_ids": [
+        "Kxd3byH9T7M_130_0to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A long, thin, crispy churro and a cup of frothy, creamy hot chocolate.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a long, thin, crispy churro being dipped into a cup of frothy, creamy hot chocolate. The churro is visibly textured and sugared, and the hot chocolate appears rich and foamy, matching the description. Additional elements like other churros on a plate and the table surface do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a long, thin, crispy churro being dipped into a cup of frothy, creamy hot chocolate. The churro is golden thin and crispyary, resembling the hot chocolate is rich and creamyamy, matching the description provided There elements like the foodurros or the wooden in a wooden setting do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ef02a947f34a5867a34313dd2a586e08ebf5c6dd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c7bf9b91b8d259ed6a80a4f087d22cc1f6842f6433d3daff41d13c770db0f9b
+size 199957
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..880df3c7e9531924672556ea092e5cdde74612f1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car, focusing on the steering wheel and dashboard. The steering wheel is black with a silver center and has a few buttons on it. The dashboard is blue and features a digital display. The car's interior is well-lit, and the steering wheel is in the center of the frame. The car appears to be a sports car, given the design of the steering wheel and dashboard. The video is likely a promotional video for the car, showcasing its interior features."
+    ],
+    "video_ids": [
+        "cBPsGaVF-kk_6_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, dashboard? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a steering wheel and dashboard as the primary focus, matching the description. The interior of the car, including the steering wheel with its emblem and the dashboard with gauges and controls, is prominently displayed. Additional elements like the gear shift and pedals are present but do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a steering wheel and a, described main focus. which the ' of The steering of the car is including the steering wheel and the logo and the dashboard with itsuges and a, is prominently displayed. The elements like the seats shift and the are also but not not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..52739a1fe5500bb2bd52cfb038f47c1702b2e79f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e206f605e4a783e0ae8f8db5a45e8adb64b22b91e1aa64325f305807ede8f4dc
+size 181742
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..fd7dc955063fcad1d06eb7229047be10e38a30ad
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and sunglasses, wearing a green jacket and a beanie. He is standing in front of a large greenhouse with a clear plastic cover. The man appears to be speaking or singing, as his mouth is open. The setting is outdoors, with trees visible in the background. The lighting suggests it might be late afternoon or early evening, as the sun is low in the sky. The overall style of the video is casual and informal, with a focus on the man and his interaction with the greenhouse."
+    ],
+    "video_ids": [
+        "OgRTYIQ4Bv8_11_354to495"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard, wearing sunglasses, a green jacket, and a beanie. His mouth is open as if he is speaking or singing.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard, wearing sunglasses, a green jacket, and a beanie, with his mouth open as if speaking or singing. The background and attire match the description, and there are no elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard, wearing sunglasses, a green jacket, and a beanie. and his mouth open as if he or singing. The description includes additional match the description, and the are no additional that contradict the given description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a11139dfafbf4ae603864f8463c06817a95ac7b9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62ac7774df142b040817d3c9280643d8b19051954b8270f4eaba97296c7e5836
+size 227561
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..88a68f66de4ea68089e52f40cb79b408306cdfba
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and glasses, wearing a cap. He is speaking and appears to be in a room with a lamp in the background. The style of the video is casual and informal, with a focus on the man's facial expressions and the context of his speech. The lighting in the room is soft and warm, creating a comfortable atmosphere. The man's attire and accessories, including his glasses and cap, suggest a relaxed and approachable demeanor. The lamp in the background adds a touch of homeliness to the scene. The overall impression is that of a personal or informal interview or discussion."
+    ],
+    "video_ids": [
+        "VdmlqdtR-E8_1_0to159"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard, wearing glasses and a cap, speaking.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard, wearing glasses and a cap, and he appears to be speaking. These elements align with the description provided. The presence of a coffee grinder in the background does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard, wearing glasses, a cap, who he appears to be speaking. The elements match with the description provided, The background of a lamp cup in the background does not contradict the core description and}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7aa39c88dbee431bf35855148be7d43f5b2e2708
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c254877862e20b46f017728a7525b205dbd60a94ca93dcd9a1ba6c4557bdb05f
+size 196409
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..0230ce8ae9da5e63ddc08ef86f1e4a7be92221a0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a white sedan car parked on a road with a scenic mountain backdrop. The car is captured from a rear three-quarter angle, highlighting its design and features. The car's taillights are illuminated, suggesting that the car is either in motion or has been recently driven. The car's design includes alloy wheels and a distinctive rear bumper. The road appears to be a two-lane highway, and the surrounding environment suggests a rural or semi-rural setting. The video is likely a promotional or review video for the car, showcasing its design and performance in a picturesque setting."
+    ],
+    "video_ids": [
+        "XiC2kbmjMWY_12_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white sedan car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a white sedan car, which is the central object. The car is clearly visible throughout the video, and its features such as the body shape, color, and wheels are consistent with a white sedan. The background elements (mountains, road, water) do not contradict the description of the object.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a white sedan car, which matches the main object in The car is depicted visible and the frames, and its design, as the ta shape, windows, and design are consistent with the sedan sedan. The background,,mountains and road, and) do not contradict the description of the car.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3d4e4bff12f9ca9e1ea30a549b42de9489515d13
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c5ca337445290b40fa4695efddbfdee384f5fd6461c551876423260da788603
+size 60035
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..a675dc65c1235e96b2644482692c2feec0ebf749
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a cooking tutorial featuring a young boy in a chef's outfit. He is seated at a table with various ingredients and a tortilla in front of him. The boy is smiling and appears to be in the process of making a dish. The table is set with bowls containing different ingredients, a jar of salsa, and a tortilla. The boy is wearing glasses and a chef's coat, suggesting that he is learning to cook or demonstrating a recipe. The setting appears to be a kitchen or a studio designed to look like a kitchen. The style of the video is casual and educational, aimed at teaching viewers how to make a specific dish. The boy's enthusiasm and the bright, clean environment contribute to the overall positive and engaging atmosphere of the video."
+    ],
+    "video_ids": [
+        "eOl8xRXU57Q_0_17to189"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young boy in a chef's outfit, bowls containing different ingredients, a jar of salsa, a tortilla, and the boy's glasses and chef's coat.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a young boy wearing a chef's coat and glasses, seated at a table with various ingredients including bowls of chopped vegetables, a tortilla, and a jar of salsa. All elements mentioned in the description are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a young boy wearing a chef's outfit and glasses, which at a table with bowls items and bowls of what vegetables, a jarilla, and a jar of salsa. The these described in the description are present and accurately depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7a04f02d65f0cf032276966db9e5a06e9cba2e52
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1284-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff9341e5a4a1ba600f935ffd6197f06eb9ae0d6f2312a411f2fa349459dce3c2
+size 167750
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..be0d4ee636969fde100194fbdab68ca0ae55770a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men sitting in the front seats of a car. The man on the left is wearing a black t-shirt and has his arm resting on the car door. He appears to be in mid-conversation, with his mouth open as if he is speaking. The man on the right is wearing a brown jacket and is looking directly at the camera with a slight smile on his face. The car is parked in a driveway with a house visible in the background. The style of the video is casual and candid, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "X-blhwYSUU8_40_54to235"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men sitting inside a car. One is in the driver's seat holding the steering wheel, and the other is in the passenger seat. Their presence and positioning match the description of 'Two men' without any contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two men sitting in a car. One is driving the driver's seat, the steering wheel, and the other is in the passenger seat. The presence and the fulfill the description of 'Two men' in any additional.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a8c0f850b919fd847f0a30f20bd0ec8347076aa5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cebeef8a0ce92d2d1c2cd591cc275e553dad8d5696e0360778d20c4332162021
+size 153545
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..6c5da94cdd966a6767b7b81e8b751b7ba54d652d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a purple thistle flower, with a focus on the intricate details of its spiky petals and the vibrant purple hue. In the first frame, a bee with a yellow and black striped body is seen approaching the flower, its wings fluttering as it prepares to land. In the second frame, the bee has successfully landed on the flower, its body nestled among the spiky petals as it begins to feed on the nectar. In the third frame, the bee is still feeding on the flower, its body still nestled among the spiky petals. The background of the video is a blurred field of tall grass, providing a natural and serene setting for the bee and the flower. The style of the video is a slow-motion, close-up shot that allows the viewer to appreciate the beauty and detail of the bee and the flower."
+    ],
+    "video_ids": [
+        "w9dfLZJloVE_2_0to173"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A purple thistle flower and a bee with a yellow and black striped body.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a purple thistle flower with two bees on it. The bees have distinct yellow and black striped bodies, matching the description. The presence of additional thistle buds and stems does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a purple thistle flower with a bees, it. One bees have yellow yellow and black striped bodies, which the description. The presence of additional beesistle flowers and green in not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cb1339245a1d2a285c1c86126d603f00757a3509
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:56cb36bb41ab08d8ece5b322c1f9085c3e5b5b72015f97e48319b4ae1717f07a
+size 242254
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..b9edf85180542e7a8ad28a86db6572ca4568e0b3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a white sports car parked in front of a building with a sign that reads \"RDB\". The car is positioned at a three-quarter angle, allowing a clear view of its rear and side. The car features black rims and a red brake light. The building in the background has a modern design with a flat roof and large windows. The sky is clear and blue, suggesting a sunny day. The style of the video is a straightforward, real-life capture with no additional effects or filters applied. The focus is on the car and the building, with no other significant objects or actions taking place."
+    ],
+    "video_ids": [
+        "OeSU5cekMQA_69_0to119"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white sports car with black rims and a red brake light.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white sports car with black rims and red brake calipers visible through the wheels, which aligns with the description. The red brake light is not clearly visible due to the angle and lighting, but the red brake calipers are present, which is a close match. The core description is largely fulfilled without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white sports car with black rims and a brake lightsipers, through the black. which aligns with the description. The car brake cal is also directly visible in to the angle of lighting, but the presence cal calipers are a, fulfilling is a part match to The car elements is largely fulfilled.\"\n contradict.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..364bd2ae7012752efc3014e29a7a04336d49bafe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b54828607268c9fc2d36893fc5e9853c7d1bc9a790592696c1c68a53544733a
+size 173343
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..0a7695f6ec4cd7ebc66a3eaf57b3bd227c785e2a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a room filled with yellow and gray electrical boxes. The boxes are arranged in rows, and the room appears to be well-lit. The style of the video is straightforward and documentary, focusing on the electrical boxes and their arrangement. The video does not contain any people or animals, and the focus is solely on the electrical boxes and their surroundings. The room itself is nondescript, with no distinctive features or decorations. The electrical boxes are the main subject of the video, and their arrangement and color are the primary elements of interest. The video does not contain any action or movement, and the focus is on the stillness and orderliness of the room and its contents. The video is likely intended to showcase the electrical boxes and their arrangement, possibly for educational or informational purposes."
+    ],
+    "video_ids": [
+        "GRiIxe3GfCs_87_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Yellow and gray electrical boxes? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features multiple yellow and gray electrical boxes arranged in rows, which matches the description. The boxes are clearly visible and occupy the majority of the visual space, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video primarily features a yellow and gray electrical boxes, in a. which align the description of The boxes are the visible and dominate the majority of the frame space, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4366430e3b1dd7d154271056ac4dbcfcc163354c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95ed24c4c0584793d011f46c8dad389cd013ea3d00c4ec17183bfad686959440
+size 61686
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..9b012fc5ee55caf50a692438d439d6dcc0e4b779
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman in a kitchen, wearing a striped shirt and a necklace. She is standing in front of a white tiled wall with a green sign that reads \"FISH POUCE\". The kitchen is equipped with a stainless steel stove and a set of knives. The woman appears to be in the middle of a cooking process, as she is holding a bottle and a spoon. The overall style of the video suggests a cooking tutorial or a cooking show. The woman's attire and the kitchen setting indicate a casual and homey atmosphere. The green sign adds a pop of color to the otherwise white and stainless steel kitchen. The woman's actions suggest that she is actively engaged in the cooking process, possibly demonstrating a recipe or technique. The presence of the knives and the bottle suggests that she is using various ingredients and tools in her cooking. The overall impression is that of a friendly and approachable cooking show, where the viewer can learn new recipes and techniques in a comfortable and familiar setting."
+    ],
+    "video_ids": [
+        "tEpRI0K0J3Q_20_0to166"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman wearing a striped shirt and a necklace, a bottle, and a spoon.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a striped shirt and a necklace, which matches the core description. A bottle (likely a green wine bottle) and a spoon (part of a utensil holder) are also visible in the background, fulfilling the 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a striped shirt and a necklace. holding matches the description description. She bottle andlikely a spice bottle bottle) and a spoon areusedially the spoonil set) are also visible in the scene, fulfilling the 'Object(s)' condition. contradict.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ff366c2abd13b4ec7f1d0701c1732aa062a6b618
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87595826cd9ca004b089aebd46f678ba5eaaa927225da52e2bfd095f3db427c1
+size 146985
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..87a64f0defc5016e9159cb9e4966fb21353812e1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is an aerial view of a rural property, showcasing its natural beauty and serene setting. The property features a large, open field with a few scattered trees, a small pond, and a white building, possibly a barn or a storage shed. The surrounding area is lush with greenery, including a dense forest and a line of trees that form a natural boundary. The sky is clear, suggesting a bright and sunny day. The overall style of the video is calm and peaceful, capturing the tranquility of the countryside."
+    ],
+    "video_ids": [
+        "XiZhuyjsKo8_95_28to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Large, open field, few scattered trees, small pond, white building (possibly a barn or storage shed).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large, open field with scattered trees, a small pond, and a white building that resembles a barn or storage shed. These elements are clearly visible and match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a large, open field with a trees, a small pond, and a white building that resembles a barn or storage shed. The elements match consistent visible and match the description provided.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e3a43a297b053ad5ff36153961b4cab5288d404b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef06c4d8de0534cf92c97bdd8dae96ab7b4d100c4f2f7f399cfc31f225fc3d18
+size 222863
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..2fbfc7d9f948771783399bfbac45b4c0876b983d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a bald man with a beard and mustache, wearing a red shirt. He is captured in three frames, each showing a different expression. In the first frame, he is looking to the left with a serious expression. In the second frame, he is looking to the right with a slight smile. In the third frame, he is looking directly at the camera with a neutral expression. The background is a plain white, which puts the focus entirely on the man. The style of the video is a simple, straightforward portrait with a focus on the man's expressions and emotions."
+    ],
+    "video_ids": [
+        "bPkJgiVb_C4_11_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bald man with a beard and mustache, wearing a red shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man with a beard and mustache, wearing a red shirt, which matches the core description. The presence of a pencil in the initial frame does not contradict the description and is an additional element that is acceptable.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man man with a beard and mustache, wearing a red shirt. which matches the description description provided The video of two plain in the background frame does not contradict the description and is not additional element that does acceptable.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c14bb5e866dad87647ee41cd855803073c1bf348
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:174df4204d8fea9b07a71738740f93b6a59853ff8a888f9e761a05d302f7a976
+size 73939
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..e85ea0e9683a16350497be87377eed4b30725624
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen enjoying a sunny day by a pool. She is dressed in a black and white polka dot swimsuit, and her hair is neatly pulled back. In her hand, she holds a green watermelon, which she seems to be admiring. The pool is surrounded by a wooden deck, and beyond it, a serene lake can be seen. The woman's relaxed posture and the tranquil setting suggest a peaceful and leisurely day spent by the pool."
+    ],
+    "video_ids": [
+        "3z5y9QllWxU_35_0to173"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a green watermelon? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a swimsuit holding a green watermelon-shaped object, which matches the core description. The object is clearly identifiable as a watermelon, and the woman is prominently featured. Additional elements like the pool and background scenery do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a woman in a swimminguit holding a green objectmelon. object, which align the description description of The presence is green green as a watermelon due and the woman is the featured in The elements like the pool and the scenery do not contradict the core.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3a30f8689a1efb1cb37633468ed81796c6a18a6e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1285-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a33168d687146d42a313ac4bc929d1ee18386ee42bf2c443d9dd0cede9142635
+size 292634
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..ea993f486f5d280de9f8a372080cca094d92db72
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two young girls sitting in a park. The first girl is wearing a blue shirt and has her hair pulled back. The second girl is wearing a white shirt and has her hair down. They are both looking at the camera with expressions of surprise or shock. The background of the video shows a grassy area with trees and a house in the distance. The style of the video is candid and informal, capturing a moment of the girls' lives."
+    ],
+    "video_ids": [
+        "RyJ6JM6_Hg0_20_205to396"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two young girls? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two young girls sitting outdoors, which matches the 'Object(s)' condition. Their facial features, hair, and clothing are consistent with being young girls, and there are no elements that contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two young girls sitting side. which ful the descriptionObject(s)' condition. The presence expressions and clothing, and clothing are consistent with the young,, and there are no additional in contradict this description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..db6ce0ef2f8cb1ce7efab1ce3b5e579475c0d8e2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e963f7493f9a63ffa15b65a95feaf27622d54366a39982b48a715fdb0a9a730
+size 110813
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..69c595eec80209a5d6a5c34079dc444d9bab7b48
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person holding a black and grey motorcycle helmet with a visor. The helmet is placed on a table, and the person is seen from the side, with their arm extended towards the helmet. The style of the video is a simple, straightforward demonstration of the helmet, likely for promotional or instructional purposes. The focus is on the helmet, with no additional context or background details provided. The lighting is even, and the image is clear, allowing for a detailed view of the helmet's design and features."
+    ],
+    "video_ids": [
+        "DnwzeJrQq4I_8_0to117"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a black and grey motorcycle helmet with a visor.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person standing behind a black and grey motorcycle helmet with a visor, which matches the core description. The helmet is prominently displayed on a table, and the person is interacting with it. Additional background elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a black's next a black and grey motorcycle helmet with a visor. which matches the description description. The helmet is prominently displayed, a wooden, and the person's partially with it, The elements elements, not conflict the main description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..af61616cbd9ce113e85f32af32a8352c363e2cb0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c6929ad07ee4cd65a079f22db057bcd2ceba8f1e7f847ad18c9597398fff049
+size 221185
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..7290078cae3c1a587a435e7d79bc606c425e4bcc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a scientific apparatus, likely used for chemical experiments or research. The apparatus consists of a clear glass flask with a long neck, connected to a clear glass tube that extends upwards. The tube is connected to a clear glass bulb, which is attached to a clear glass tube that extends downwards. The apparatus is mounted on a metal stand with a blue clamp. The background is black, which contrasts with the clear glass of the apparatus. The style of the video is educational, likely used in a laboratory or classroom setting to demonstrate the use of the apparatus."
+    ],
+    "video_ids": [
+        "8GVSuKkuLzY_12_186to368"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A clear glass flask with a long neck, a clear glass tube extending upwards, a clear glass bulb, another clear glass tube extending downwards, and a metal stand with a blue clamp.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a clear glass apparatus that includes a flask with a long neck, a vertical glass tube extending upwards, a bulbous section (glass bulb), and another tube extending downwards. It is also supported by a metal stand with a blue clamp, matching the described elements. Additional tubing and connectors do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a clear glass flask that includes a flask with a long neck, a clear tube tube, upwards, a bulbous flask,which bulb), a a tube extending downwards. The is mounted supported by a metal stand with a blue clamp, which the description elements. The elements and a are not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..137b085d77ef3c22a7ed3acda8592504574fc9b6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8d21700b98bb30fb1d9d4ac1a0ffd9c01bb1b9a8c8436de568fa41145ee418b0
+size 61581
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..395ae0007d618989d7080338c1e9cd36cf50098e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of disappointment and reflection for a group of football players. The players, dressed in white and blue uniforms with the number 24 prominently displayed, are seen walking off the field. One player is seen wiping his face with his hand, possibly in frustration or sadness. The other players are seen looking down, their heads bowed in defeat. The background is a blur of green, suggesting a stadium filled with spectators. The overall mood of the video is somber and reflective, capturing the aftermath of a lost game."
+    ],
+    "video_ids": [
+        "Js68KWJkSvw_19_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Football players dressed in white and blue uniforms with the number 24 prominently displayed.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows football players in white and blue uniforms, and the player on the left clearly has the number 24 prominently displayed on his jersey. The description is largely accurate, with no major contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows football players wearing white and blue uniforms with with the number in the right has has the number 24 prominently displayed on his jersey. The players is largely matched, and the significant contradictions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..90d012954c50aa9996ff005e499374e80d6f130a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d29e1e37e798f32d809c2f3cfe34bf4013e5e7909c75c48bd02230c65750f186
+size 286969
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..6d40ef03fd8d09464bddd1b5db58d3b479527daa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a gray and white cat being wrapped in a beige towel. The cat appears to be comfortable and relaxed as it is being swaddled. The cat's eyes are open and it seems to be looking directly at the camera. The towel is being wrapped around the cat's body, covering it completely. The cat's ears are sticking out from the top of the towel, and its whiskers are visible. The cat's tail is not visible in the frame. The cat's fur is soft and fluffy, and it appears to be well-groomed. The towel is soft and plush, and it seems to be providing a warm and cozy environment for the cat. The cat's eyes are bright and alert, and it seems to be enjoying the attention it is receiving. The cat's mouth is closed, and it does not appear to be making any sounds. The cat's nose is visible, and it seems to be breathing normally. The cat's eyes are a light color, and they are looking directly at the camera. The cat's ears are perked up, and it seems to be listening attentively. The cat's whiskers are white, and they are sticking out from the sides of its face. The cat's tail is not visible in the frame. The cat's fur is a mix of gray and white, and it appears to be well-groomed. The towel is a"
+    ],
+    "video_ids": [
+        "8e0AWKlxgBU_14_1186to1360"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ['gray and white cat', 'beige towel']? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a gray and white cat wrapped in a beige towel, lying on a white surface. The cat's fur pattern and the towel's color match the description. There are no conflicting elements that contradict the specified objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a gray and white cat, in a beige towel. which on a textured surface. The cat's fur and and the towel's color match the description provided There are no additional elements in contradict the given objects.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8f67dd6eac4da7a45451bbfab92c15cf21faa47f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0e45ed0c27bde622a1154047a8c73527a3f423f6f786ebcf8c043098420e335
+size 349716
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..7c7029221d226d84002ff8931ccd469ea87bd419
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a woman with blonde hair sitting at a table with a white sewing machine. She is wearing a striped shirt and has a smile on her face. She is holding a spool of thread in her hand and appears to be in the process of sewing. The table is covered with various colorful fabrics and buttons. The background features a white wall with yellow polka dots. The style of the video is bright and cheerful, with a focus on the woman's enjoyment of sewing."
+    ],
+    "video_ids": [
+        "BR4t23BCYJ0_41_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a white sewing machine, a spool of thread, colorful fabrics, buttons? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman sitting at a table with a white sewing machine in front of her. She is handling colorful fabric pieces (likely felt or paper circles) and appears to be sewing them. Although a spool of thread is not clearly visible, the context implies it's part of the sewing setup. The colorful fabrics and buttons are represented by the various colored circular pieces scattered on the table. The scene matches the described objects without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman sitting at a table with a white sewing machine, front of her. She is holding a fabrics and andincluding sp) similar)) and a to be preparing them. There the spool of thread is not visible visible, the presence suggests its is present of the sewing setup. The presence fabrics and buttons are also by the felt fabric felt pieces and on the table, The presence is the description objects without any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..880cb42cf3514493ce69a62bf3087163eecefc55
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:09bf2e565d971423675c155b68a26562259a960f8596602ce042604b1fb39559
+size 191403
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..fa129078946b58478fa2194a983ebe47ce04fcb5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a stop-motion animation featuring three dolls set against a green backdrop with white flowers. In the first frame, a person's hand is holding a small doll with a red bow in its hair. In the second frame, the same doll is now standing next to a larger doll wearing a white dress with a black bow. In the third frame, the larger doll is now standing next to a male doll wearing a white tuxedo. The dolls are positioned in a way that suggests a narrative, with the male doll looking at the female doll. The style of the video is playful and whimsical, with a focus on the dolls and their interactions."
+    ],
+    "video_ids": [
+        "XfUI2uc9eu8_9_280to425"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ['small doll with a red bow', 'larger doll wearing a white dress with a black bow', 'male doll wearing a white tuxedo']? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a small doll with a red bow (held in hand), a larger doll wearing a white dress with a black bow (centered in the frame), and a male doll wearing a white tuxedo (in the background). All three specified objects are present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showing showing a small doll with a red bow,top by the), a larger doll wearing a white dress with a black bow,standinged in the frame), and a male doll wearing a white tuxedo (standing the bottom). The objects objects objects are present and match the descriptions provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..52d3d47caa8ffa4daf86cb4d70aa336a875e3442
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:65cbc5a48a39a5405ea2a6760e5405e819b4926e861c50c97221f51461dbfc46
+size 91853
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..18c2189603be5cc646af6ece723d9a73b8dc3202
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a thrilling moment at a snowboarding event. The crowd of spectators, bundled up in winter gear, watches in anticipation as the snowboarder takes off from the ramp. The snowboarder, dressed in a vibrant red jacket, is captured mid-air, performing an impressive trick. The snowboarder's body is angled towards the ground, suggesting a high level of skill and control. The snow-covered slope and the clear blue sky in the background add to the excitement of the scene. The video is shot from a high angle, providing a bird's eye view of the action, and the camera follows the snowboarder's trajectory, creating a sense of motion and excitement. The overall style of the video is dynamic and action-packed, capturing the adrenaline-fueled spirit of the sport."
+    ],
+    "video_ids": [
+        "AaZ5grifocc_0_0to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Snowboarder, crowd of spectators? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a snowboarder descending a snow-covered ramp and a large crowd of spectators watching the event. These elements align with the 'Object(s)' condition described, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a snowboarder performing a slope-covered slope, performing crowd crowd of spectators watching the event. The elements directly with the 'Object(s)' condition,, making there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..eca9c11f59273645ad98fac7be4c523c18047ffb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1286-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9576ea00e2adf43cd32dc52ccc3dfdb39006cd363666864a715870f007b3d880
+size 294507
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..7739f61f05b81c72ba709b4fd01e7e9e6797c5e0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen sitting in a cozy living room, her smile radiating warmth. She's dressed in a vibrant green shirt, adding a pop of color to the scene. The room around her is filled with personal touches - a fireplace adorned with various knick-knacks, a comfortable couch, and a chair. The walls are decorated with framed pictures, each telling its own story. The overall atmosphere is one of comfort and homeliness, with the woman as the central figure, inviting viewers into her personal space."
+    ],
+    "video_ids": [
+        "CdpkZTn1a5o_21_0to112"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman sitting in the living room, a vibrant green shirt, a fireplace, various knick-knacks, a comfortable couch, a chair, and framed pictures.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a woman sitting in a living room, wearing a vibrant green shirt. Behind her is a fireplace with various knick-knacks on its mantel. A comfortable couch and a chair are visible in the room, along with framed pictures on the walls. All core elements described are present and consistent with the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a woman sitting in a living room. wearing a vibrant green shirt. The her, a fireplace with various knick-knacks on the mantel. There comfortable couch and a chair are also, the background, and with framed pictures on the wall. The elements elements of in present and accurately with the video depicted}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..38a7020240f010119c66d49983634f8aecc9af07
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:08ee1b9823fc62cc1423ce6c9f7f1d0af82f5d6f15e412d7ac7c302dd252dfbb
+size 97407
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..64a70805c038beaf932d2a9f26f64275a27d5579
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a cooking process, specifically the boiling of a liquid in a pot. The pot is placed on a stove, and a thermometer is inserted into the liquid to monitor its temperature. The liquid appears to be a syrup or a sauce, as it has a thick consistency. The stove is set to a high heat, and the temperature of the liquid is rising. The kitchen is well-lit, and there are various kitchen utensils and items in the background, including a cutting board, a bowl, and a potted plant. The style of the video is a straightforward, real-life cooking demonstration, with no additional effects or embellishments. The focus is on the cooking process and the use of the thermometer to ensure the liquid reaches the desired temperature."
+    ],
+    "video_ids": [
+        "335IGVWOAkg_34_0to134"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A pot, a thermometer, a liquid (syrup or sauce).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a pot filled with a bubbling liquid, which appears to be oil suitable for frying, and a thermometer is inserted into the liquid to monitor temperature. Although the liquid is not explicitly syrup or sauce, it is consistent with the context of cooking (frying) and the thermometer's presence supports the object description. The core elements (pot, thermometer, liquid) are present and correctly depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a pot on with a yellowbling yellow, which appears to be a or for frying. as a thermometer being being into the oil. measure its. The the liquid is not explicitly described or sauce, it is a with the context of cooking,potrying), which the presence's use suggests the condition condition of The scene elements ofpot, thermometer, liquid) are present and match interpreted.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cd5c7faa0cfa678a2a9e2bd4760d7233c7a5e094
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d75cc58b08f010d4e642859eb2b248c4f94d3ca770a8a5424be8fcf7ebd57c4b
+size 121292
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..b68db1d6bc04155fdd8deb108a17074709dbe60f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a simple, educational animation that illustrates the nutritional value of various fruits and vegetables. The style is hand-drawn and chalkboard-like, with each frame featuring a different food item and its corresponding calorie count. The foods include broccoli, red onions, cucumbers, carrots, and tomatoes. The calorie counts are written in chalk next to each food item, providing a visual representation of the nutritional content of these healthy choices. The overall tone of the video is informative and encourages viewers to consider the nutritional benefits of incorporating these fruits and vegetables into their diets."
+    ],
+    "video_ids": [
+        "mof2vs71plA_7_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Broccoli, red onions, cucumbers, carrots, and tomatoes? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully displays broccoli, red onions, cucumbers, carrots, and tomatoes as part of a nutritional chart on a chalkboard. Each vegetable is clearly visible with its corresponding calorie count, fulfilling the 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video contains ful images, red onions, cucumbers, carrots, and tomatoes, part of the list chart. a chalkboard. The vegetable is clearly labeled and its name name count, which the 'Object(s)' condition.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0cb89ccd4d0dcb8c09d5d492f34e7869d36c6f0e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0d6987bac7e941e7b80cc005eaa2b2aa403c9d884199d88da026d77a1deb1234
+size 82909
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..6efa6b6b3e873ee80e6384b5bd1f8dfc2e119a64
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a jubilant moment at a football game. A football player, wearing a green and yellow uniform, is the center of attention. He is surrounded by a group of people, all dressed in similar attire, indicating they are fans or possibly teammates. The player is holding a football, suggesting he has just scored a touchdown or made a significant play. The crowd is cheering and celebrating, with some people even lifting the player off the ground in a show of excitement and support. The atmosphere is one of triumph and camaraderie, as the player and his supporters share in the joy of the moment."
+    ],
+    "video_ids": [
+        "X6Ec-hZaToA_64_97to233"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A football player in a green and yellow uniform, a crowd of people, a football.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a football player in a green and yellow uniform (Green Bay Packers), a crowd of enthusiastic fans, and a football being held by the player. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a football player in a green and yellow uniform,number Bay Packers), a crowd of people people, and a football being held by the player. The elements elements of in present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b51e547235cd8bd89b6e492650af1ccfd20ee3cf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f0cc5c69b2d287d1cd6f8db208cfdfeda260fd6503de86c7a2f9580eb3d02798
+size 303081
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..636ef00c37baf28820ecfa3b43058a4770be63a7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a silver sports car with a sleek design, set against a dark background. The car is positioned in the center of the frame, with its front facing the camera. The car's design includes a low and wide stance, with a long hood and short rear deck. The wheels are large and appear to be made of a shiny material, possibly chrome or polished metal. The car's body is smooth and shiny, reflecting the light in the room. The car's headlights are off, and the windows are tinted, making it difficult to see inside. The car's license plate reads \"M-WD 7260\". The overall style of the video is sleek and modern, with a focus on the car's design and details. The dark background helps to highlight the car, making it the main subject of the video. The video does not contain any text or additional objects. The car is the only object in the video, and it is stationary. The video does not show any movement or action. The car is the only object in the video, and it is stationary. The video does not show any movement or action. The car is the only object in the video, and it is stationary. The video does not show any movement or action. The car is the only object in the video, and it is stationary. The video does not show any movement or action. The car is the only object in the video, and it"
+    ],
+    "video_ids": [
+        "bLyWH6_-ijM_58_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Silver sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a silver sports car, specifically a BMW, which matches the 'Object(s)' condition. Additional text overlays (e.g., 'FRANKFURT', 'AUTO', 'SHOW', 'SEPTEMBER', '1989') do not contradict the core description of the object, as they provide context rather than altering the primary subject.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a silver sports car, which a sleek M which align the descriptionSilver(s)' condition. The elements and ande.g., 'BMWONTLINURT') 'M', 'M') '2TEMBER') '26')9') are not contradict the core description of the object but as they are context about than altering the main focus of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..eec7b07f86edf99e6a6843f510affeacbc74ddeb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b14ae1dddfaa75c08fdece043fe7e09b4c959db38f26b154cbe92f478269d665
+size 116578
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..5f7ab4b93d0cab2437f2a24df526538a7bdcb41e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two young men are in a kitchen, preparing and eating food. The first man, wearing a black t-shirt with white text, is standing on the left side of the frame. He is looking at the second man, who is wearing an orange tank top with white text. The second man is holding a piece of food in his hand and is about to eat it. In front of them is a wooden dining table with a blue bowl on it. The bowl contains food, and there are also some bottles and a knife on the table. The kitchen has white cabinets and a window in the background. The overall style of the video is casual and informal, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "I0wuhB8mP2A_25_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two young men, a wooden dining table, a blue bowl, bottles, a knife.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two young men sitting at a wooden dining table. A blue bowl is visible on the table, and bottles can be seen in the background on shelves. A knife is also present on the table near the bowl. All specified objects are present and do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two young men standing at a wooden dining table. There blue bowl is present in the table, and there and be seen in the background. the. A knife is also present on the table. the bowl. The the objects are present and match not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f7def26071299c7f501675cf63cc66ecb60dda40
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d796549097c4628ac33d606947145eccf3a25ffa822b4f4687728744c662fd91
+size 91667
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..8a87796ce87699e6c80280886bb393979cda042f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a tranquil scene of four black pigs in a grassy field. The pigs are engaged in various activities, with one of them drinking from a white and black water jug. The jug is placed on a red plastic tray, which is positioned on the ground. The pigs are scattered around the tray, with some of them standing close to it, while others are a bit further away. The grassy field provides a natural backdrop to this scene, with the green of the grass contrasting nicely with the black color of the pigs. The overall style of the video is simple and straightforward, focusing on the pigs and their interaction with the water jug."
+    ],
+    "video_ids": [
+        "BUzxClvIs04_12_0to217"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Four black pigs, a white and black water jug, a red plastic tray? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows four black pigs gathered around a white and black water jug placed on a red plastic tray, which matches the described objects. The pigs are actively interacting with the area around the jug, and the background elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows four black pig, around a red and black water jug placed on a red plastic tray. which matches the description objects. The scene are interacting interacting with the jug, the jug, and the setting consists, not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f0eef3cf071db979370b914cdec4f06f287e36fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2729143410ece851459e04884c894e134c89700894092c246d2f872372050d70
+size 203310
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..64db2b1f822dfe8ed573efb84fc30e9e18969e24
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the dynamic movement of a bright orange sports car on a track. The car's sleek design and vibrant color make it the focal point of the video. The car is seen from a low angle, emphasizing its speed and power as it races along the track. The background is filled with blurred lights, indicating the car's high speed and the excitement of the race. The car's headlights are on, illuminating the track ahead and adding to the sense of motion. The video is a thrilling depiction of a high-speed race, with the orange sports car as the star of the show."
+    ],
+    "video_ids": [
+        "b_4cMiIn7po_25_0to122"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bright orange sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a bright orange sports car, specifically a McLaren, with clear details of its front fascia, headlights, and wheels. The car is the central focus and matches the description perfectly.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a bright orange sports car, which a Porsche, which its details such its design designia and headlights, and aer. The car is depicted central focus, matches the description of.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c383f0d318bf4fb8bb9c23122334fc340f1dc663
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1287-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d6c4e64049d478a4b38922f05c7708f58573a0ac4cb1d38a2b80b41bacf4ebf0
+size 212706
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..c3ecc4dfa8d8ae2ea506029e2e957b3f29a1892d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a basketball player in a red jersey with a white headband. He has a beard and mustache, and his expression is intense. The player is looking off to the side, possibly focusing on the game or his teammates. The background is blurred, but it appears to be a basketball court with a crowd of spectators. The style of the video is a close-up shot, capturing the player's facial expression and intensity. The lighting is bright, highlighting the player's features and the red of his jersey. The overall mood of the video is competitive and focused."
+    ],
+    "video_ids": [
+        "2L6mW0BRR04_6_168to313"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Basketball player in a red jersey with a white headband, beard, and mustache.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a basketball player wearing a red jersey, a white headband, and has a visible beard and mustache, matching the core description. The background and other elements do not contradict this depiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a basketball player wearing a red jersey with which white headband, and has a beard beard and mustache. which the description description provided The background appears additional elements are not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1683cd6f11541853f3f3ea24ad5e4154df1a6017
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9aa4f5e5a7e87ef5859f21d39ee3fc426bf91dac0ce3c641eaf6a916668b71f0
+size 160862
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..63db5198f84a5babab9859e7fd7fd16ee9b917e1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a thrilling moment of a Peugeot rally car racing through a muddy track. The car, adorned with the Total and Red Bull logos, is in motion, kicking up a spray of water and mud as it speeds along. The driver, focused and determined, navigates the challenging terrain with skill and precision. The car's headlights pierce through the spray, illuminating the path ahead. The surrounding environment is a blur of green, indicating the car's high speed and the intensity of the race. The video is a dynamic and exciting portrayal of the adrenaline-fueled world of rally racing."
+    ],
+    "video_ids": [
+        "fdKNtNYY5D8_4_228to414"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ['Peugeot rally car', 'Driver', 'Headlights']? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a Peugeot rally car with visible headlights, and although the driver is not distinctly visible inside the cockpit, the car's features and branding match the description. The presence of additional elements like water splashes and sponsor logos does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows depicts a rallyugeot rally car, a headlights and which the the car is not directly visible, the car, the presence's design and the are the description. The presence of a elements like the andashes and a logos does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7362b21c7ff400818d8a8f4d52870055066b1c31
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ae4d8bb4f6c1263c3220d0e7195ed80cfd5f9575be187ce4f331e8f6aa565da
+size 307888
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..ae807e237ade69084c10899e808986af795a94cb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a bicycle's rear wheel and drivetrain, focusing on the chain, gears, and spokes. The bicycle is a modern road bike with a sleek frame and a black color scheme. The chain is in motion, indicating that the bicycle is being pedaled. The gears are clearly visible, showing the bike's ability to shift through different speeds. The spokes of the wheel are also visible, demonstrating the bike's structural integrity. The background is blurred, but it appears to be an indoor setting, possibly a garage or a workshop. The style of the video is a time-lapse or slow-motion shot, capturing the intricate details of the bicycle's components in a visually appealing manner."
+    ],
+    "video_ids": [
+        "H87rRqECYqg_16_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bicycle's rear wheel, chain, gears, spokes? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the bicycle's rear wheel, chain, gears, and spokes as requested. The rear wheel is prominently displayed, the chain connects the crankset to the cassette, the gears (cassette and derailleur) are visible, and the spokes are clearly seen connecting the rim to the hub. All elements are accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a rear's rear wheel, chain, and, and spokes. described. The image wheel is visible featured, with chain is the gearsset to the rear, and gears arecassette) chainleur) are visible, and the spokes of part seen radi the wheel to the hub. The these are consistent represented without any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..59b5fe22c8d5717eed738f5481fdee3d49c7c50a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:856d67eb3474eb3111dc9d18bc98c94b41d46466d3a33a3ee3a8c22eefc47899
+size 179130
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..e5aaa4acde1a125742147ef51141a674b059dce0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment from MSNBC, featuring a man standing outdoors in front of a building. The man is wearing a checkered shirt and has a microphone clipped to his shirt, indicating that he is a reporter or news anchor. The background shows a clear sky and a building with a green awning. The text overlay on the video reads \"MUELLER REPORT IMPACT\" and \"KEY SWING STATE VOTERS WEIGH IN ON MUELLER FINDINGS\", suggesting that the segment is discussing the impact of the Mueller report on voters in key swing states. The style of the video is typical of news broadcasts, with a focus on delivering information in a clear and concise manner."
+    ],
+    "video_ids": [
+        "HQFhF42wYeM_2_69to194"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man wearing a checkered shirt with a microphone clipped to his shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a checkered shirt with a microphone clipped to his shirt, which matches the description. The background and additional elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a checkered shirt, a microphone clipped to his shirt. which matches the description provided The background appears additional elements do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8b8500e21ca41dc754e06dc61156448e4203bd76
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:93d4d8d2a6630f9437fe3a1a33459a8d7553a57283ff7c682ffda3638b249703
+size 269486
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..7dff20d4b5400673929cd828e738adeaa492502c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a body of water, likely a lake or pond, viewed from an aerial perspective. The surface of the water is calm, reflecting the surrounding trees and sky, creating a mirror-like effect. The trees on the right side of the frame are dense and dark green, indicating they might be coniferous. On the left side, there are trees with lighter, possibly autumn-colored foliage, suggesting a change in seasons. A small, indistinct object, possibly a bird or a floating item, moves slightly across the water's surface, adding a subtle dynamic element to the otherwise tranquil setting. The overall atmosphere is peaceful, with the stillness of the water and the quiet presence of nature dominating the scene. There are no significant changes in the environment throughout the video; it maintains its calm and reflective quality, emphasizing the natural beauty and tranquility of the location."
+    ],
+    "video_ids": [
+        "2830c98b04a6f3e13cf5615682167020526614ab9bd64635f985f798df3947e9"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Small, indistinct object moving across the water's surface.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small, indistinct object moving slowly across the water's surface, which aligns with the 'Object(s)' condition. The object is visible as a faint white speck against the misty, reflective water, and its movement is consistent throughout the frames. The surrounding reflections of trees and mist do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a serene, indistinct object moving across across the surface's surface. which aligns with the 'Object(s)' condition. The object appears barely as a slight, speck in the darky reflection reflective water, consistent it movement is subtle with the frames, The rest environment and trees and sky add not obscure the observation observation.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f53e138b8ebcfdc280a238245945832c08ace527
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aec74beb01dc12d7fabc46e9a143cd3e3e7db6f136a93b636df5f0495c163dda
+size 124295
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..0501ab847bc6b1b1ad2e0165a96b4bbf74d3686a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dolphin performing a trick in a pool. The dolphin is partially submerged in the water, with its head and back visible above the surface. It is holding a hula hoop around its head, which is a common trick performed by dolphins in marine shows. The hula hoop is yellow and green striped, adding a vibrant contrast to the blue water. The dolphin's skin is smooth and gray, and it appears to be focused on maintaining its grip on the hoop. The water around the dolphin is slightly disturbed, indicating recent movement. The background is a blurred view of the pool's edge, suggesting an indoor or controlled environment typical of a marine park or aquarium. The lighting is bright, highlighting the dolphin and the hoop against the darker water. The overall scene conveys a sense of playful interaction between the dolphin and its trainer, showcasing the animal's intelligence and agility."
+    ],
+    "video_ids": [
+        "6699c291a12af068d0d6a380767048928436806e8154b1731044ace248b720ec"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A dolphin and a yellow and green striped hula hoop.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a dolphin with a yellow and green striped hula hoop balanced on its head while partially submerged in water. The core objects described are present and accurately depicted without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a dolphin in a yellow and green striped hula hoop around on its head, in submerged in water. This presence elements, in present, match depicted, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..97eccbe972cf5fdbb2c7e08067f6d39c617fcfff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0dba17ad0e8eaa2d5dc9db855265d48d7cdcdbadf416d37340d8a9d7849fb474
+size 339575
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..61fd784d896b61915c487d20fecb411840c96d47
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young boy standing in a lush, green forest. He is wearing a white t-shirt with a colorful design on the front. The boy is looking directly at the camera with a surprised expression on his face. The forest around him is filled with large, vibrant green leaves and pink flowers. The boy appears to be in the middle of the forest, surrounded by nature. The overall style of the video is natural and serene, capturing the beauty of the forest and the boy's reaction to it."
+    ],
+    "video_ids": [
+        "1a3BLPQC9X8_3_0to158"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young boy in a white t-shirt with a colorful design, looking directly at the camera with a surprised expression.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young boy wearing a white t-shirt with a colorful design, and he is looking directly at the camera with a surprised expression. The surrounding foliage does not contradict the core description and is consistent with the setting.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young boy wearing a white t-shirt with a colorful design, looking he is looking directly at the camera with a surprised expression. The background environment and not contradict the description description of is an with the setting of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..75ccb1e36902634597141dc15816f2b6441f03e8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a2cbc58ba0b388d93812638d10071d7ceffe1a1afb9e4615d68d0a66990ab192
+size 171955
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..22578306ce4515cdc95694b8382f9de5a2527dd1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features an older man with a mustache and glasses, standing in front of a vintage car. He is wearing a gray jacket and appears to be speaking or presenting. The car is a classic model, possibly a convertible, and is parked on a dirt road. The setting appears to be an outdoor area with a brick building in the background. The man's expression is serious, and he seems to be addressing the camera or an audience. The overall style of the video suggests it could be a documentary or a news segment, focusing on the man and the car."
+    ],
+    "video_ids": [
+        "3B0XGs1fdrk_23_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An older man with a mustache and glasses, wearing a gray jacket, and a vintage car (possibly a convertible).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an older man with a mustache and glasses, wearing a gray jacket, standing in front of a vintage car (which appears to be a light green convertible). The description matches the core elements of the video without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an older man with a mustache and glasses, wearing a gray jacket, standing in front of a vintage car.possibly appears to be a convertible blue convertible). The setting matches the core elements of the video, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..28c3bfd6b6f27b70989753ae0461f192062413d2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1288-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0ba4284cdae58700962d252909c60b2d4605e8aa97ba2bdfc7674d124772773f
+size 144149
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..2b513264950939cc7778f784d8bc7174f5461282
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen enjoying a ride in a light blue vintage car. The car is parked in a parking lot, surrounded by other vehicles. The woman, dressed in a black tank top and sunglasses, is standing on the roof of the car with her arms outstretched, embracing the moment. The car is a classic model, with a white license plate and a blue logo on the side. The scene captures a moment of joy and freedom, as the woman takes in the surroundings from her unique vantage point."
+    ],
+    "video_ids": [
+        "Qo6_uUdYrjU_29_0to133"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A light blue vintage car, a woman, other vehicles in the parking lot.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by prominently featuring a light blue vintage car, a woman standing on its roof, and other vehicles in the background parking lot. All described elements are clearly visible and consistent with the provided description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting featuring a light blue vintage car, a woman standing behind the hood, and other vehicles in the background, lot. The elements elements are present visible and match with the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0107d0b2b6e766d76d66942a86bd869ad93de88a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b655dbd5e3f9a1adb22979345264f6232edaa3c000e8b0fda863cbd98dc284e
+size 200673
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..775305b88e98c855386e7313be9d466248c7a334
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of two drinks placed on a wooden table. The first drink is a glass filled with a yellow liquid, possibly a cocktail, garnished with a sprig of mint. The second drink is a pineapple cut in half, filled with the same yellow liquid and garnished with a sprig of mint. The drinks are placed side by side, and the camera angle is slightly elevated, providing a clear view of the drinks and their garnishes. The style of the video is simple and straightforward, focusing on the presentation of the drinks without any additional context or action. The lighting is bright and even, highlighting the colors of the drinks and the garnishes. The wooden table provides a natural and rustic backdrop for the drinks, enhancing their visual appeal."
+    ],
+    "video_ids": [
+        "I26zTL9Dgvc_11_1097to1233"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ['glass filled with yellow liquid (possibly a cocktail)', 'pineapple cut in half', 'sprig of mint']? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a glass filled with yellow liquid (likely a cocktail), a pineapple cut in half and used as a drink container, and sprigs of mint as garnish. These elements are clearly visible and match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a glass filled with a liquid,likely a cocktail) a pineapple cut in half, used as a cup holder, and aigs of mint garn garnish. These elements match consistent present and match the description provided.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..59650aceb62cbdadf0e6599ce35e4baab7e72e08
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6fdf93735ae52e0b085c56284767cb756b05708aed38aa9782093d752134ad6
+size 85435
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..bd421954c82c2bb7b9c200c361630ea3b7f0f65b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene underwater scene featuring two silver fish swimming in a tank. The fish, with their shimmering scales, are the main focus of the video. They are seen swimming in a sandy area, their bodies glistening as they move through the water. The tank itself is filled with various plants and rocks, providing a naturalistic environment for the fish. The video is shot from a high angle, allowing for a comprehensive view of the tank and its inhabitants. The overall style of the video is calm and peaceful, capturing the beauty of these underwater creatures in their artificial habitat."
+    ],
+    "video_ids": [
+        "hIqazW2dAI0_32_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two silver fish? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two silver fish with detailed scales and black-tipped fins, consistent with the description. Their actions and environment do not contradict the core description, and additional elements (like other fish or plants) are present but do not conflict with the primary focus on the two silver fish.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows shows a silver fish swimming a scales and a stripesipped fins, which with the description of The appearance and the, not contradict the ' description.\"\n and the elements likelike the fish or objects) do not but do not conflict with the main focus on the two silver fish.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..305b6135d2e0de5a910b0971b1c9e245a04f5227
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:919444a9361afaeffb69be7d83225686397e2a63b006e68688f94a2d609b22a3
+size 251169
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..7a2d26a12503ada5675db8c30954001278a9be10
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a restaurant, enjoying a meal. He is seated at a table, holding a piece of food in his hand, and appears to be in the process of taking a bite. The man is wearing a pink shirt and has a bald head. The restaurant has a menu board with Japanese characters, suggesting that the cuisine is likely Japanese. The lighting in the restaurant is warm and inviting, and there are other patrons visible in the background. The style of the video is casual and candid, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "VJBpPvfmvgE_13_117to318"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a table, a piece of food, other patrons? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man holding and eating a piece of food, with other patrons visible in the background. While a table is not prominently featured, its presence is implied by the setting of a restaurant. The core elements described are all present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man in a eating a piece of food, which a patrons visible in the background. The the table is not explicitly featured, it presence is implied by the setting, the restaurant. The man elements of in present present in contradiction.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bcfa02566adfc6fb0e220a1dacad8a9680071fff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0936dcf480b0cb8c99ab88db7fe8ca71504df73abc0e915d3b080017c0290ea3
+size 139454
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..4e726efd0f57bac75c2f9702b13743b2f30eee5e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men sitting at a table in a casual setting, possibly a restaurant or a cafe. The man on the left is wearing a blue shirt and is looking at the man on the right, who is wearing a yellow shirt and glasses. The man on the right is holding a fork and appears to be eating or about to eat. On the table, there are three white cups and a plate with food. The background features red chairs and a white table, suggesting a modern and clean interior design. The style of the video is candid and informal, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "kRyFzx7PCVM_12_153to304"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men sitting at a table; one in a blue shirt, the other in a yellow shirt with glasses. There are three white cups and a plate with food on the table.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men sitting at a table, one wearing a blue checkered shirt and the other wearing a yellow shirt with glasses, which matches the description. There are three white cups and a plate with food on the table, also matching the description. The only deviation is the appearance of a woman in a green dress in the last frame, but this does not contradict the core description, as the primary focus remains on the two men and the objects on the table.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men sitting at a table, one in a blue shirtered shirt and the other wearing a yellow shirt with glasses. which matches the description. There are three white cups on a plate with food on the table, which align the description. The presence additional is that presence of a fork in the yellow shirt in the background frame, but this does not contradict the core description and as the question focus remains on the two men and the table on the table.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..aa55f61d664d62949607b2527cf85ffe08bb0784
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c5061052c063d0e41131d30d97757aa48b1f9a51748b017448d3d74e00cf8caa
+size 110095
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..a5d64f3ae2c87eb95752d8efea8b6aba022fb210
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the dynamic interaction between the ocean and a small island. The first frame shows the island, lush with greenery, surrounded by the vast expanse of the ocean. The second frame shows the ocean's surface, with waves crashing against the island's shore, creating a frothy white foam. The third frame shows the ocean's surface, with the waves receding, leaving behind a trail of white foam. The video is shot from a high angle, providing a bird's eye view of the island and the ocean. The style of the video is realistic, capturing the natural beauty of the scene without any artificial embellishments."
+    ],
+    "video_ids": [
+        "_CpX_08aYKY_41_0to110"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Island, Ocean? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows an island covered with green vegetation and a surrounding ocean with visible waves. These elements match the 'Island, Ocean' condition described, with no conflicting elements present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts an island with in greenery and surrounded sandy ocean with waves waves. The elements directly the 'Island' Ocean' condition.. indicating the additional additional present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..872410df2e93dffa9466d6c693502126d53668fd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e1363ea64e4d8adbc8d146ad9eebe0aa87e5711133e60a21b13e0620bc4a4606
+size 277253
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..80c38e84c18fcc0083b89a9ca13894163d2f123f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment on a soccer field. A male soccer player, dressed in a white jersey with red accents and red shorts, is in motion. He is running towards the right side of the frame, his body leaning forward in anticipation. His arms are outstretched, ready to receive a pass or make a move. The background is a blur of pink and red, suggesting a stadium filled with spectators. The focus is solely on the player, emphasizing his speed and agility. The video is a freeze-frame of a high-energy moment, capturing the intensity and excitement of the sport."
+    ],
+    "video_ids": [
+        "5whpKIp4hlo_29_91to302"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A male soccer player in a white jersey with red accents and shorts.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a male soccer player wearing a white jersey with red accents and red shorts, which matches the description. The jersey features the Red Bull logo and Nike branding, and the player's attire is consistent with a professional soccer uniform. The background is blurred, but it does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a male soccer player wearing a white jersey with red accents and shorts shorts, which matches the description provided The player has a number Bull logo, the branding, which the player is attire is consistent with a soccer soccer uniform. The background, a but but it appears not contradict the description description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..827c7d85bcd0b9194d51d00c21fca1d5bf27868e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5496a6a65d962f13959d89e002f244d72632fec9e9c6029b9d5338f762a9663
+size 281001
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..1749505ef19d7f14892e15a40983d02693f64790
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of intense focus and concentration from a male soccer player. He is wearing a blue jersey, and his face is set in a serious expression. His eyes are focused on something off-camera, and his mouth is slightly open as if he is about to speak or shout. His hand is raised to his face, with his fingers gently touching his cheek, suggesting a moment of deep thought or strategizing. The background is blurred, but it appears to be a soccer stadium, indicating that this is a professional match. The style of the video is a close-up shot, focusing on the player's face and upper body, which emphasizes his intense expression and the importance of the moment."
+    ],
+    "video_ids": [
+        "yZwDurMzS0I_7_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A male soccer player wearing a blue jersey? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a male soccer player wearing a blue jersey, which matches the description. The jersey has visible details such as a logo on the sleeve, and the player's appearance and attire are consistent with a professional soccer player. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a male soccer player wearing a blue jersey, which matches the description. The player is a team such as a logo and the shoulder, and the player is attire, attire are consistent with a soccer soccer player. The background suggests setting elements in not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0c41cd8efeb0a362b411bff4b6ab7698d859fa15
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1289-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:55765f50d0b2121e8cc001e6094b920d26b5dde57e7c8f13aed12ed1d8268c82
+size 158863
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..c3207ecdd73da93284ff2775dc3b11993001d692
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a step-by-step tutorial on how to make a lemon and honey drink. The style of the video is simple and straightforward, focusing on the ingredients and the process of making the drink. The first frame shows a wooden cutting board with a lemon, a lime, and a ginger root. The second frame shows a glass filled with water, lemon slices, and a sprig of mint. The third frame shows a jar of honey with a wooden spoon. The video likely includes voiceover instructions on how to prepare the drink, emphasizing the use of fresh ingredients and the natural sweetness of honey. The overall tone of the video is clean and healthy, suggesting that the drink is a refreshing and nutritious option."
+    ],
+    "video_ids": [
+        "7vxROgUOf88_31_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Lemon, lime, ginger root, glass filled with water and lemon slices, sprig of mint, jar of honey, wooden spoon? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a glass of tea with lemon slices and mint, a jar of honey with a wooden spoon, and also includes lime halves and ginger root as described. All specified items are present and accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showcasing showing a glass filled water with lemon slices, a leaves a jar of honey, a spr spoon, and a includes le,, a root in part. The the objects are present and in depicted in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..88a1f07b9febdb59f0c3267ffd6b18f70430b540
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:362b69ea521ebd3539c7f529b6a5378ccd100fd01553a2b2bdc5d75bb512edaa
+size 159540
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..76aaa2756724769032679b664d21f4aaa5636292
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a blue shirt sitting under a tree. He appears to be in a relaxed state, with his eyes closed and mouth slightly open. The man is leaning against the tree, suggesting a casual and comfortable setting. The background is blurred, but it seems to be an outdoor environment with natural light. The style of the video is candid and informal, capturing a moment of rest or contemplation."
+    ],
+    "video_ids": [
+        "AMkyMNlSHqw_2_0to119"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a blue shirt, sitting under a tree, leaning against it? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue shirt sitting on the grass under a tree, with his body leaning against the tree trunk. This matches the core description provided. The presence of a woman sitting nearby does not contradict the description, as the focus is on the man and his position relative to the tree.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue shirt, under a ground, a tree, leaning his head leaning against the tree trunk. The matches the description description of, The presence of the camouflage in on does not contradict the main, as it focus is on the man in his position under to the tree.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..836fd584d28b17f6f6ad60de3d24397b0855b894
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd4c18e8490ba8113b2ef873387a05a9c5730e415a3cad21a7bb425e720cb657
+size 140494
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..dbf6cd19e83e9e494a530535288df2f9728fc24d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person's hand reaching into a bowl filled with cherry tomatoes. The hand is in the process of picking up a tomato. The bowl is placed on a table, and there are other bowls and plates of food nearby. The food includes a variety of vegetables such as peppers and greens. The style of the video is a close-up shot, focusing on the hand and the bowl of tomatoes. The lighting is bright, highlighting the colors of the tomatoes and the other vegetables. The overall atmosphere of the video is casual and homey, suggesting a scene of someone preparing a meal."
+    ],
+    "video_ids": [
+        "YKw7J9JPB0M_8_0to164"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person's hand, bowl of cherry tomatoes, other bowls and plates of food, vegetables (peppers and greens).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person's hand interacting with a bowl of cherry tomatoes, alongside other bowls and plates containing food items like artichokes and greens, as well as vegetables such as peppers and leafy greens. All elements described in the condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a person's hand interacting with a bowl of cherry tomatoes. which other bowls and plates of food and such greenichokes and green. as well as vegetables such as green. greensy greens. The these described in the condition are present and do depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..095710da127cdf3d89a46abb1a89a41aba271777
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c1ee1ab05f3fee5d0dbca38e226feb67203e106cc05c1aacd3e03be28a015074
+size 178042
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..103e1549526ff9b40501ad634b6a1566c8e66828
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in front of an American flag. He is dressed in a suit and tie, and appears to be speaking or listening intently. The man is the main subject of the video, and his position in front of the flag suggests that he may be a public figure or an official. The flag serves as a backdrop, adding a patriotic element to the scene. The style of the video is straightforward and professional, with a focus on the man and his interaction with the flag. The video does not contain any additional objects or elements, and the man's actions are the primary focus."
+    ],
+    "video_ids": [
+        "LCEOP0YjSgQ_10_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man dressed in a suit and tie, an American flag? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed in a suit and tie, standing in front of an American flag, which matches the described 'Object(s)' condition. The man is clearly identifiable as Barack Obama, and the American flag is prominently displayed behind him. The presence of a news channel watermark and date overlay does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed in a suit and tie standing standing in front of an American flag. which ful the description 'Object(s)' condition. The presence is wearing wearing as a Obama, and the presence flag is prominently displayed behind him. There presence of the flag ticker logo in a in does not contradict the core description but}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..60dc1b5b6b493198db1480985b5a8c266a734246
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:20bc9ef6f4f48d6604f493a7ef633dedf542a46fd13a106874c717ff5af5478e
+size 132298
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..b1f956a2d7957641b25365e8526bcbe7410ad41b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse of a car journey on a winding mountain road. The car, a white SUV, is seen driving along the road, which is surrounded by lush green trees and grass. The road curves and twists, following the contours of the landscape. In the distance, majestic mountains rise against the clear blue sky, adding a sense of scale and grandeur to the scene. The time-lapse effect creates a sense of motion and progression, as the car moves along the road, the landscape changing subtly with each passing moment. The overall style of the video is serene and peaceful, capturing the beauty of nature and the joy of a road trip."
+    ],
+    "video_ids": [
+        "otIrj-VugbA_49_20to156"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white SUV? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white SUV driving along a winding road through a natural landscape with mountains in the background. The vehicle is clearly visible and matches the description of a white SUV, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white vehicle driving on a winding road. a scenic landscape. mountains in the background. The vehicle is the visible and matches the description of a white SUV. fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3476eae58ca86876542d568b2f642628a0814d2c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:372561c53d6f591f4886da3887b431a80e9a03642bc50f59daee68b07e3adc81
+size 322693
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..c7b9c92cb55081c8534cccc497cad5dde4afe15f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is on two players, one in a white uniform and the other in a black uniform. The player in the white uniform is in possession of the football, holding it securely in his hands. He is in the process of running, his body leaning forward in anticipation of the play. The player in the black uniform is in pursuit, his body language suggesting he is trying to tackle the player with the ball. The background is filled with other players, their uniforms a mix of colors, indicating they are part of the same team. The field they are playing on is a vibrant green, contrasting with the players' uniforms. The video is a freeze-frame of a high-energy moment, capturing the intensity and excitement of the game."
+    ],
+    "video_ids": [
+        "EOcqsfgx_Ag_17_0to110"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two players - one in a white uniform with the ball, another in a black uniform pursuing him.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two primary players: one in a white uniform holding the football and another in a black uniform actively pursuing him. These elements align with the specified condition, even though other players and background elements are visible, they do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts two players players: one in a white uniform holding a football and another in a black uniform pursuing pursuing him. The elements match with the ' '. making though there players and details elements are present, they do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b84cd797cf74d22b586af6d5788673bea0ebc51b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37a14fe1992a00e62c9ce061ad7c02dec05ad001694391053ed1b0e809c01206
+size 284532
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..3f8d76c97849c232007cc04c547a74980f27d05c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a woman in a store. She is wearing a blue polka dot dress and has her hair up. The store has a red and gold color scheme, with mannequins and clothing on display. The woman is looking to the side, and there is a text overlay that reads \"\uc544\uc774\ub4e4 \uc0c1\uc810\uc5d0\uc11c \uc5b4\ub5bb\uac8c \uc0b4\uc544\uac00\ub294\uc9c0?\" which translates to \"How do children live in the store?\" The video is likely discussing the experiences of children who live in a store, possibly as part of a social experiment or documentary. The style of the video is informative and journalistic, with a focus on the woman and her surroundings."
+    ],
+    "video_ids": [
+        "aH254NujBVM_20_52to234"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a blue polka dot dress with her hair up? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The woman in the video is wearing a blue dress with white polka dots, and her hair is styled up in a bun. These details match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video in the video is wearing a blue pol with white polka dots, and her hair is styled up in a bun. The elements match the description provided. the questionObject(s)' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..14038c45a18c8c6c95fb20ab976c492749ff6a10
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5f73c1ec6f0596ce7535e090efd249f6dff48b0b3006d44aaf89fd5a382d66a
+size 186604
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..5fc58fabca54bbc98efb3380ddd5c5fe7779ee6b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a blue pickup truck parked on a grassy field. The truck is a modern model with a chrome grille and silver rims. The truck is parked on a grassy field with a wooden fence in the background. The sky is overcast and the grass is green. The truck is the main focus of the video and is shown in three different angles. The style of the video is a straightforward, real-life depiction with no special effects or animation. The video is likely meant to showcase the truck's design and features."
+    ],
+    "video_ids": [
+        "0OmDq5zJ9-0_8_0to190"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue pickup truck? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a blue pickup truck, specifically a GMC Denali model, parked on grass. The vehicle's color, type, and branding are consistent with the description. Additional elements like the background scenery do not contradict the core object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a blue pickup truck, which a Ford Sierraali,, which in a. The truck is design, design, and design are consistent with the description of There elements like the grass and do not contradict the core description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4c971d5c1c5c8690e9be433cc3f44f95f3e499b6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-129-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76ddd5b5fbc961f6cfabe1aa6ca98f8eb2ad8db41bd305febd3d2b33946dcd18
+size 193730
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..d045e6f19636a2ad9cfbd9f181d4993008109a53
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a brown bear is seen in a forest setting. The bear is holding a bowl in its paw, and it appears to be eating something from the bowl. The bear is standing in front of a tree, and there is a teapot nearby. The bear's fur is brown, and it has a black nose. The forest setting is lush and green, with trees and foliage surrounding the bear. The bear seems to be enjoying its meal, and the overall atmosphere of the video is peaceful and serene."
+    ],
+    "video_ids": [
+        "rKYEzUQaQ2s_34_43to210"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Brown bear, bowl, teapot? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a brown bear holding a bowl and drinking from it. A teapot is also visible in the background on the right side of the frame. All three specified objects are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a brown bear holding a bowl and a from it, There teapot is also visible on the background, the ground side of the frame. The the elements objects ( present in the with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3f9a58904eaea11ec3d010bd622209575abf7abf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:160190cc96c049db98afcafc7f8b2bfa17906ac879c2b180b5ee4f5e60956c2d
+size 260799
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..d68beb35e12af3bad0450b304a29aba290643ae4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with glasses and a white cap, who is looking to the side. He is wearing a white hoodie and has a beard. The background consists of a colorful, reflective surface that resembles a CD pattern. The man appears to be in a casual setting, possibly indoors. The style of the video is candid and informal, capturing a moment in the man's life."
+    ],
+    "video_ids": [
+        "bBzw5pH4g_M_31_0to145"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with glasses, a white cap, a white hoodie, and a beard, looking to the side.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing glasses, a white cap, a white hoodie, and has a beard, and he is looking to the side. These elements match the description provided in the 'Object(s)' condition. The background of CDs does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing glasses, a white cap, and white hoodie, and a a beard. looking he is looking to the side. The elements match the description provided, the questionObject(s)' condition.\"\n The background, colorful does not contradict the description description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c02efc511de3cdb93fd125fa6e58607a2e5fd388
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1371f5c942f330f6337b7b7a81caaa46e1845bcd6af6e7e35a6628842cb2e5f6
+size 109591
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..95ca1bae5e9c135630bac2478af7b0f8f6064364
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the process of a pancake being flipped with a fork. The pancake is golden brown and appears to be freshly cooked. The fork is silver and has four tines. The pancake is being flipped over onto a black plate. The plate is placed on a white countertop. The background is blurred, but it appears to be a kitchen setting. The style of the video is a close-up shot with a shallow depth of field, focusing on the pancake and the fork."
+    ],
+    "video_ids": [
+        "MLYYYkTAFtk_40_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Pancake, fork, plate? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a pancake being cooked on a griddle, with a fork used to flip it. A plate is visible in the background, fulfilling the 'Object(s)' condition as described. The presence of additional elements like steam and oil does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a pancake being interact on a plateiddle, with a fork being to lift the. The plate is also in the background, and the 'Object(s)' condition. it. The presence of syrup elements like syrup and syrup does not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..12798093d30fd72c46074fd9b6c71e6a787c9b1a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59d55fb29105b0cda31d7a14880850781975564e6deac51082b719dbcd44a7e1
+size 221682
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..e2d466f10f91dca04d704d6cfa91fc36e7fba214
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated scene featuring three anthropomorphic rabbits standing on a grassy bank by a body of water. The rabbits are dressed in clothing and accessories, with one wearing a green hat, another in a blue jacket, and the third in a pink dress. They are engaged in a conversation, with one rabbit gesturing towards the water. The background shows a serene landscape with trees and a calm river, suggesting a peaceful, natural setting. The style of the animation is colorful and detailed, with a focus on the characters and their expressions."
+    ],
+    "video_ids": [
+        "2krd47cN2ko_0_0to131"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three anthropomorphic rabbits, one with a green hat, another in a blue jacket, and the third in a pink dress.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows three anthropomorphic rabbits. One rabbit is wearing a green hat with a red pom-pom, another is dressed in a blue jacket, and the third is wearing a pink dress. These characters are the central focus and match the description exactly.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts three anthropomorphic rabbits. One is is wearing a green hat, a blue backpack-pom, another is in in a blue jacket, and the third is wearing a pink dress. The elements match standing main focus of fulfill the description provided.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0c07cc7d01ec374e80ba64d9215f5030989eeff2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b7ecfcf909025e0dc443b42ada33f11c3828fad92bc18d8bdfa0fcedd1100f73
+size 135145
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..a139c6cbaf84c14067142d85548027db5b6a4007
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a teal-colored Volkswagen car parked on a gravel surface. The car is positioned at a slight angle to the camera, allowing a clear view of its front and side. The car features a distinctive black stripe running along its side, and a roof rack is installed on top. The license plate of the car reads \"WOB-PR 11\". The car appears to be in good condition, with no visible damage or wear. The background of the video is a cloudy sky, suggesting that the weather is overcast. The gravel surface on which the car is parked is uneven, with small rocks scattered about. The overall style of the video is straightforward and unembellished, focusing solely on the car and its immediate surroundings."
+    ],
+    "video_ids": [
+        "G1LfyBtx86k_25_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A teal-colored Volkswagen car with a black stripe and a roof rack. The car has the license plate 'WOB-PR 11'.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a teal-colored Volkswagen station wagon with a black stripe along the lower body and a roof rack. The license plate is clearly visible as 'WOB-PR 11'. All key elements of the description are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a teal-colored Volkswagen car wagon with a black stripe and the side side and a roof rack. The license plate visible visible visible as 'WOB-PR 11'. The these elements of the description are present, match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4da07fc89e51c7091f670f94662b4364085e6065
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:674fea7c5d6c8dee0cecee8aa2a48d2d8c3313e77923724b2750a30fe68d1ec0
+size 270556
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..8f06c012dca36606f8c59b644dd6c00be6457410
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman in a kitchen, standing in front of a refrigerator. She is wearing a purple shirt and has blonde hair. The kitchen is well-equipped with various appliances and utensils. The refrigerator is large and metallic, and there are several bottles on the counter. The woman appears to be in the middle of a conversation or demonstration, as she is gesturing with her hands. The overall style of the video is casual and informative, likely aimed at providing cooking or kitchen tips."
+    ],
+    "video_ids": [
+        "2ERNLRnUj_0_36_0to170"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a purple shirt with blonde hair, a refrigerator, and several bottles on the counter.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair wearing a purple shirt, standing in a kitchen with a refrigerator visible in the background. Several bottles can also be seen on the counter behind her. These elements match the core description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in blonde hair wearing a purple shirt, standing in front kitchen with a refrigerator in in the background. There bottles are be be seen on the counter in her. The elements match the description description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5b1abe1d2209ae94f8dd2e66bf0ee7a682be0bf0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:16a89f59c7226dddf7e90ef0d0fa19e8bbe30fc8269d246975a770bd478b5fc7
+size 178638
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..ea68205f807ec727630b4a3f4789a120a774672a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young man with glasses and a beard, standing in a kitchen with a window in the background. He is wearing a plaid shirt and appears to be speaking. The style of the video is a casual, personal interview or testimonial. The man is the main subject of the video, and the kitchen setting suggests a homey, comfortable atmosphere. The window in the background provides natural light, which illuminates the man and the kitchen. The overall tone of the video is relaxed and informal."
+    ],
+    "video_ids": [
+        "ZbH1k-wk7v0_1_0to128"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man with glasses and a beard, wearing a plaid shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young man with glasses and a beard, wearing a plaid shirt, which matches the core description. Additional elements such as text overlays and background items do not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man with glasses and a beard, wearing a plaid shirt. which matches the description description provided The elements such as the or or background noise are not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5151c841a749a119eee6c6b8cce204b2e02fa0a0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68e2065ed3930d56d8d5ac7473902c05d28e0aff51bb1c40dfb805b9b3882db9
+size 119689
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..be127a79e1236ddb74fa3bc404b3ba84b47f618a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man sitting in a studio, speaking into a microphone. He is wearing a white hoodie and a jacket with a floral pattern. The studio has a cityscape in the background, and there is a logo on the wall that reads \"DIS\". The man appears to be engaged in a conversation, possibly a radio interview or a podcast recording. The overall style of the video is professional and polished, with a focus on the man and his interaction with the microphone."
+    ],
+    "video_ids": [
+        "7n3ETLy1r28_9_0to181"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a microphone, a white hoodie, a jacket with a floral pattern, a cityscape, a logo reading 'DIS'.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting in front of a microphone, wearing a white hoodie under a jacket with a floral pattern (roses). The background features a cityscape, and the logo 'DISH NATION' is visible, which includes 'DIS' as requested. All core elements of the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man wearing in front of a microphone, wearing a white hoodie and a jacket with a floral pattern.whiches and The background includes a logoscape and and there logo readingDIS''ATION' is visible, which includes theDIS'. in part. The elements elements of the description are present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2a38ba602bce449bf667ebba87538c3068971677
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1290-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb1027877783371baf8267b396af31b7c55d5ac9e2346900fdfe5bc28fa6a4ed
+size 211425
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..7dae717349e243e67442b4fa4b969f802eae8b20
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a car's tire in motion. The tire, with its shiny silver hubcap and black rubber, is the main focus of the video. The car is blue, and the tire is positioned on the side of the car, giving a sense of depth and perspective. The tire is in motion, creating a blur effect that adds a sense of speed and movement to the video. The background is blurred, drawing the viewer's attention to the tire and the car. The video is shot from a low angle, looking up at the tire, which adds to the dramatic effect. The overall style of the video is dynamic and energetic, capturing the essence of motion and speed."
+    ],
+    "video_ids": [
+        "D5B0d2kePh4_20_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A car's tire with a shiny silver hubcap and black rubber.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a car's tire with a shiny silver hubcap and black rubber, which matches the description. The tire is rotating, but this does not contradict the core description of the object. The blue car body and grated floor are additional elements that do not conflict with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a car's tire with a shiny silver hubcap and black rubber. which matches the description provided The tire is in, and the does not contradict the core description of the object. The presence car body and the road are additional elements, do not conflict with the main of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5fb8086b65664a5933b0eb06bdbc1be6d9a94fa7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17084b1dc658ec0ebde2f2b50478bee9beb4c359abc6df028f045d53cb60653d
+size 226016
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..07ea998017e7cb67f596049b8ec9672b91c03a37
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the picturesque view of a medieval town perched on a hill, with a castle overlooking the landscape. The town is nestled amidst lush greenery, with the castle standing tall in the center. The sky above is a clear blue, dotted with fluffy white clouds. The sun casts a warm glow on the scene, highlighting the vibrant colors of the town and the castle. The video is taken from a high vantage point, providing a panoramic view of the town and the surrounding landscape. The castle, with its imposing structure, stands as a testament to the town's rich history. The town itself is a charming blend of old and new, with cobblestone streets and modern buildings co-existing harmoniously. The video beautifully captures the essence of the town, its history, and its natural beauty."
+    ],
+    "video_ids": [
+        "Cu-Pxwe1rj4_11_0to102"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Medieval town, castle, lush greenery, sky, clouds, sun? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by showcasing a medieval town perched on a hill, surrounded by lush greenery, with a prominent castle-like structure (likely a cathedral or fortress) at its center. The sky is visible with dynamic clouds, and the warm lighting suggests the presence of the sun, even if not directly visible. All core elements described are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting a medieval town withched on a hill, featuring by lush greenery. with a clear castle in structure.likely the church or church) at the center. The sky is depicted with clouds clouds, and the lighting lighting suggests the presence of the sun, even though the directly visible in The elements elements of in present and contribute with the scene.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5179181321f00af46641cabbdfce57d984646bc1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ab80e84dbfdd69f188e5dc41d7416aea4a65ed5a9d06747044e1f8c72fc0da8
+size 162939
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..0627d83f5441b1fb24da15bdd45d3bf3444244e9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases an intricate arrangement of dominoes on a wooden surface, forming various shapes and patterns. The scene begins with a close-up view of blue dominoes meticulously arranged to create a detailed design resembling a bird or a similar figure. As the camera pans across the table, it reveals more of the elaborate setup, including yellow dominoes forming a large, intricate pattern that appears to be a stylized representation of a bird's wings or feathers. The camera continues to move, highlighting additional sections of the design, which include green and white dominoes contributing to the overall complexity and beauty of the artwork. The background features a wall made of black and yellow dominoes, adding depth and contrast to the scene. Throughout the video, the camera smoothly transitions from one section of the domino arrangement to another, providing a comprehensive view of the entire setup. The final frames show the completed designs, emphasizing the skill and creativity involved in creating such a detailed and visually striking piece using domino"
+    ],
+    "video_ids": [
+        "ad07a27c5e7f2036a887447c575db5a1311d00a72c3af5cd181648d8f2c1213d"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dominoes (blue, yellow, green, and white)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features dominoes arranged in various colors including blue, yellow, green, and white, forming intricate patterns and shapes. These dominoes are the central focus of the video, fulfilling the 'Object(s)' condition as described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features blueinoes in in a colors, blue, yellow, green, and white. which a patterns and structures. The elementsinoes are the central focus of the video, and the 'Object(s)' condition as described.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..49d37b8766bd2a72abd1b909738994abad742e53
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8cfd53bb9a9a498a824a7a80bbc7454d2aa73497304a978e096192d4a2c6be84
+size 276103
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..ebf4255259d5013283f35fe137d3b81fa490c36d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a gray and white cat being wrapped in a beige towel. The cat appears to be comfortable and relaxed as it is being swaddled. The cat's eyes are open and it seems to be looking directly at the camera. The towel is being wrapped around the cat's body, covering it completely. The cat's ears are sticking out from the top of the towel, and its whiskers are visible. The cat's tail is not visible in the frame. The cat's fur is soft and fluffy, and it appears to be well-groomed. The towel is soft and plush, and it seems to be providing a warm and cozy environment for the cat. The cat's eyes are bright and alert, and it seems to be enjoying the attention it is receiving. The cat's mouth is closed, and it does not appear to be making any sounds. The cat's nose is visible, and it seems to be breathing normally. The cat's eyes are a light color, and they are looking directly at the camera. The cat's ears are perked up, and it seems to be listening attentively. The cat's whiskers are white, and they are sticking out from the sides of its face. The cat's tail is not visible in the frame. The cat's fur is a mix of gray and white, and it appears to be well-groomed. The towel is a"
+    ],
+    "video_ids": [
+        "8e0AWKlxgBU_14_1186to1360"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ['gray and white cat', 'beige towel']? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a gray and white cat wrapped in a beige towel, lying on a white surface. The cat's fur pattern and the towel's color match the description. There are no conflicting elements that contradict the specified objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a gray and white cat, in a beige towel. which down a surface surface. The cat's fur and and the beige's color match the description provided There are no additional elements in contradict the given objects.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..11ad67ef4c4121498a316653e988badfe3d1aa0c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:875476c263e91e9d71c5a85a93b99d1417e5f4f4e40ce392d804e4539ffd8298
+size 202444
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..aced94dd2da82f8cbb6230be1b01a2600e0c6c1b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated depiction of a samurai warrior in a traditional Japanese setting. The samurai is shown in three different poses, each showcasing a different aspect of his attire and weaponry. In the first frame, the samurai is seen from the front, wearing a gold and red armor with a helmet adorned with two long horns. He holds a sword in his right hand, which is raised in a defensive stance. In the second frame, the samurai is shown from the side, revealing the intricate details of his armor and the sword he carries. In the third frame, the samurai is seen from the back, holding the sword in his left hand, ready for battle. The background features a blurred image of a traditional Japanese landscape, adding to the overall atmosphere of the video. The style of the video is realistic with a focus on historical accuracy, capturing the essence of the samurai warrior in his traditional attire and setting."
+    ],
+    "video_ids": [
+        "PV9od98uYEE_7_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A samurai warrior? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video displays two figures dressed in traditional samurai armor, complete with helmets featuring horn-like protrusions, red face paint, and swords. These elements strongly identify them as samurai warriors, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a sam dressed in traditional samurai armor, which with helmets, horns-like protrusions, which and paint, and a. The elements are align the as samurai warriors, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..96b8004ed28dfa09e9d29f6b06edd3d51128131d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50c461031db1af47510263241540f907b7453c7e94c1b32fa2614e3c1e55cc66
+size 384950
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..d5eb2e1a94e0b93ac74d80893d240b391c7c5ec1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a kitchen, preparing food. He is wearing a black baseball cap and a green t-shirt. The kitchen is equipped with wooden cabinets and a black microwave. On the counter, there are several bowls and a spoon. The man is focused on his task, and the overall atmosphere of the video is casual and homey."
+    ],
+    "video_ids": [
+        "L1jMH39dMqE_47_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black baseball cap and a green t-shirt, several bowls, and a spoon.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black baseball cap and a green t-shirt, standing in a kitchen. Several bowls and a spoon are visible on the counter behind him, matching the described objects. The core elements of the description are accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black baseball cap and a green t-shirt, which in a kitchen. He bowls and a spoon are visible on the counter. him. and the description elements. The man elements of the description are accurately represented in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..77ca774cd55121d8bd2f9e6b191ef32623658a90
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e9459e8a4506ba01676a8f104a7ff3e95359215866dd2d973a2bf6ba62373f28
+size 138218
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..8581df978e45067eddc46b0da4c5fda8d7c08267
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is an aerial view of a lush green property with a variety of buildings and vegetation. The property is surrounded by a dense forest, and a road runs alongside it. The buildings are scattered across the property, with some closer to the road and others further back. The vegetation is a mix of trees and shrubs, providing a vibrant green backdrop to the buildings. The road is lined with a grassy median, and there are a few cars driving on it. The overall style of the video is a bird's eye view, providing a comprehensive look at the property and its surroundings."
+    ],
+    "video_ids": [
+        "I9Ihm34kAFg_11_120to262"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Buildings, vegetation (trees and shrubs), road, cars? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing buildings with varied roof colors, dense vegetation including trees and shrubs, a road with visible cars, and surrounding greenery. All specified elements are present and accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing buildings, blue roof colors, vegetation vegetation including trees and shrubs, a road running a lanes, and a naturalery. The elements elements are present and contribute depicted in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..66ee5e66d1d9aca7eea1b75b48bcffb2450dd6e8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e8f8bc89b1bebd0a34bd13170779c71e4adfa2d081c73266b7e70dbed90c669
+size 175228
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..8c446dbea48df1f068bdb0b65d7a467eb3cf4d23
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a meal being enjoyed. The meal consists of a sandwich, a side of salad, and a small bowl of sauce. The sandwich, which is the main focus of the video, is made with a slice of bread and is topped with melted cheese and a dollop of sauce. The salad, which is placed to the side of the sandwich, is a mix of greens and other vegetables. The small bowl of sauce, which is placed next to the sandwich, is filled with a dark-colored sauce. The video is shot in a realistic style, with the camera focusing on the details of the meal. The colors in the video are vibrant and the lighting is bright, highlighting the textures and colors of the food. The video does not contain any text or narration."
+    ],
+    "video_ids": [
+        "iIdqebAN3W0_14_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Sandwich (main focus), salad, small bowl of sauce? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a sandwich as the main focus, topped with a poached egg and hollandaise sauce, accompanied by a side salad and a small portion of dark sauce on the plate. These elements align with the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a sandwich as the main focus, with with a sauceached egg and aaise sauce, which by a side salad and a small bowl of sauce sauce in the side. The elements clearly with the description of,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..34a6c6924b3eef48b2b8b38ebe9064ce7dafe746
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1291-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:737b050db4fe61214fccc96932573b79110083930cc8d880b98b4a34371516c4
+size 98690
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..61a57113e367501a6a3c1f6fc5ed6c41ae6eef12
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man adjusting his black headphones. He is wearing a blue shirt with a patch on the left side. The man is standing in a park-like setting with trees in the background. The style of the video is casual and informal, capturing a moment of everyday life. The focus is on the man and his actions, with the background serving as a simple, unobtrusive setting. The video does not contain any text or additional elements."
+    ],
+    "video_ids": [
+        "FkuDaqf2S48_6_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man wearing a blue shirt with a patch on the left side, adjusting black headphones.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue shirt with a visible patch on the left side, and he is adjusting black headphones around his neck. These elements match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue shirt with a patch patch on the left side, and he is adjusting black headphones. his head. The elements match the description provided, the questionObject(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e2ae122bf72ebff98a8d6dacabf780fdeb7d817e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3b74391485a8a98aae556feb42a5809593d48471d493602c54be9154f63bb71
+size 229782
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..859424c35a81d563d95c81452afc517af71a381f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man with long hair is standing in a room with a black wall and a gray striped curtain. He is wearing a black and gray jacket and holding a green object in his hands. The man appears to be in a relaxed posture, with his hands clasped together. The room has a modern and minimalist design, with a large screen and a desk visible in the background. The overall style of the video is casual and informal, with a focus on the man and his actions."
+    ],
+    "video_ids": [
+        "IbYG7q9OO18_20_46to190"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with long hair, wearing a black and gray jacket, holding a green object, and standing in a relaxed posture with his hands clasped together.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with long hair wearing a black and gray jacket. He is holding a green object (which appears to be a circuit board design on his shirt, not a separate object). He is standing with his hands clasped together in a relaxed posture, which matches the description. Additional elements in the background do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with long hair wearing a black and gray jacket, He is holding a green object inpossibly appears to be a small board or) a jacket). not a physical object) He is standing in his hands clasped together in a relaxed posture. which matches the description. The elements like the background, not conflict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..898f702a4fd1ecdc0b2834d64523a6ab9abaa5ee
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:465606e6a760a5420ca24d62a16a5e28cfa7c6da6fef0176b2c62bb31fa10edf
+size 81788
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..14960c9ac0f119ebcdae7feb0661cb1f975c78c1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a car being lifted by a hydraulic lift in a garage. The car, a sleek black model, is positioned on the lift, its front end facing the camera. The lift is yellow and blue, providing a stark contrast to the car's dark exterior. The garage itself is filled with various tools and equipment, indicative of a well-equipped workspace. In the background, a person can be seen, possibly the mechanic working on the car. The video is a snapshot of a typical day in a garage, with the car being serviced and the mechanic going about his work."
+    ],
+    "video_ids": [
+        "JNH-8U6S54s_48_0to122"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A sleek black car and a yellow and blue hydraulic lift? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a sleek black car elevated by a yellow and blue hydraulic lift, which matches the core description. Additional elements like other vehicles and tools in the background do not contradict this primary focus.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a sleek black car elevated on a yellow and blue hydraulic lift, which matches the description description provided The elements like the cars and equipment in the background do not contradict the main focus.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a80d51954b38ad2cd6e6adece79ce2e29c47bad2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff3728b3254c83cbac6a6a5293a027dd206b749ed92f6a61c30d9f95b4589ee3
+size 176514
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..ea2798a68ea170143a28ce658c9afa0877bec6f3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic scene at a race track. The first frame shows a car speeding down the track, its tires hugging the curve as it rounds a bend. The second frame shows the car continuing its journey, now further down the track, with the landscape stretching out behind it. The third frame shows the car approaching a chicane, a series of tight turns that require precise handling. The track itself is surrounded by a dry, arid landscape, with a few trees scattered here and there. The sky above is clear and blue, suggesting a perfect day for racing. The style of the video is a high-angle shot, providing a bird's-eye view of the action on the track. This perspective allows for a comprehensive view of the car's movement and the surrounding landscape."
+    ],
+    "video_ids": [
+        "EmD6pDS6nLk_7_0to194"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a car driving on a racetrack, fulfilling the 'Object(s): A car' condition. The car is visible throughout the sequence, navigating the track, and there are no elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a car driving on a racetrack. which the 'Object(s)' A car' condition. The presence is the in the frames, and the curves, which there are no elements that contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..34c86d60d3947526739b1f3c1950e1e52271c810
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d690e412d70cd57c9909cb883aac5a1614cb8b25aad9e5383c742c93c02b313
+size 192490
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..f6b4ccb503fded5f4515edfa552234ee558e9438
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a silver BMW car driving down a country road. The car is sleek and modern, with a distinctive design that includes a large rear window and a prominent rear bumper. The car's license plate reads \"BZM-533\". The road is surrounded by lush green trees and a clear blue sky. The car is moving at a moderate speed, and the driver appears to be focused on the road ahead. The overall style of the video is realistic and naturalistic, capturing the beauty of the countryside and the elegance of the BMW car."
+    ],
+    "video_ids": [
+        "V9TvBS7GdYA_26_157to323"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A silver BMW car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a silver BMW car from the rear, with visible BMW branding, model designation, and license plate. The car's design, color, and details are consistent with a silver BMW, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a silver BMW car driving the rear view which the details branding and a features, and a plate. The car is design and including, and features align consistent with the BMW BMW, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0c3a0a26f507a593ca3b57ce707845a4be99e5b2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7ce4e61be571ac824f012dc88e5fbc4ae714b944c923e4a5d53c3c713ede3ed2
+size 294563
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..5f5473704835d94eae2277c581c532531fffdf7e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a BMW car, taken from the perspective of the passenger side door. The car's interior is well-lit, with the sun shining through the windows, casting a warm glow on the seats and dashboard. The steering wheel, which is the focal point of the image, is prominently displayed in the center of the frame. The dashboard, which is visible in the background, features a digital display and various controls. The car's seats are upholstered in a light-colored material, and the door panel is equipped with a window control and a side mirror. The overall style of the video is clean and modern, with a focus on the car's interior design and features."
+    ],
+    "video_ids": [
+        "U6zRNxhzvJM_50_53to182"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, dashboard, seats, door panel, window control, side mirror? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the steering wheel, dashboard, seats, door panel, window controls, and side mirror as requested. All these elements are visible and accurately represented within the interior view of the car, matching the description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows the interior wheel, dashboard, seats, door panel, and control, and side mirror, part. The these elements are visible and match depicted in the frame of of the car. fulfilling the ' provided any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5433969a164c67a65844044f527285b153045f24
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bad9475a39f55a60eeecf3e43e93df321cc9ee729869a8e7a83ce06634744771
+size 169277
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..239159e567d2c9524d208452504e99ad6d4b1834
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man with glasses and a red jacket standing in front of a brick wall with a window. He appears to be speaking or gesturing with his hands. The window reflects the image of another person, possibly a woman, who is also visible in the background. The man's expression is serious, and he seems to be engaged in a conversation or presentation. The style of the video is candid and informal, capturing a moment in the man's day."
+    ],
+    "video_ids": [
+        "4iZE-xqsAks_19_0to162"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a brick wall, a window, and a reflection of another person.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a red jacket and glasses standing in front of a brick wall and a window. The window reflects another person, fulfilling all the specified conditions. Additional elements (like posters or reflections of other people) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man wearing a red jacket, glasses, in front of a window wall. a window. The reflection reflects another person, which the the conditions conditions. The elements likelike the or other) other people) are not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4c7286b4b86f9afd3f415504b608f9429789d7dc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ebd13f9a3d9fc0261c88a705e9631a09be288e2e81016d9b7dea9fcd86ff86f9
+size 167091
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..ea8dfebb6b804b1630f2b3428152339b8da1aed3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a bowl of food on a marble countertop. The bowl contains a creamy white substance, possibly a type of soup or porridge, topped with sliced hard-boiled eggs and chopped green onions. The food is garnished with toasted nuts, adding a crunchy texture to the dish. The style of the video is simple and straightforward, focusing on the food without any additional context or background. The lighting is bright and even, highlighting the textures and colors of the ingredients. The video does not contain any text or narration, allowing the viewer to focus solely on the food. The overall impression is one of a healthy and nutritious meal, possibly served as a breakfast or brunch dish."
+    ],
+    "video_ids": [
+        "cYhA5_RJRXY_1_0to194"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bowl of creamy white substance, sliced hard-boiled eggs, chopped green onions, toasted nuts.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bowl containing a creamy white substance, topped with sliced hard-boiled eggs, chopped green onions, and what appears to be toasted nuts or garlic. These elements match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bowl containing a creamy white substance, which with sliced hard-boiled eggs, chopped green onions, and toasted appears to be toasted nuts. seeds. The elements match the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ba89854bb124e6f6ce4af499b934c06865146aec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1292-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:910321b329d997421695dd2fbc17cd09ed5c49469624ef5f346053ad47df325a
+size 83270
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..e0a78ece7ec1fe30e0993467d1d0af61b1112c0a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a large, white, curved greenhouse with a series of plants growing inside. The plants are neatly arranged in rows, and the greenhouse appears to be well-maintained. The structure of the greenhouse is visible, with a series of metal supports and a white covering that allows sunlight to pass through. The greenhouse is situated in a rural area, with trees and a clear blue sky in the background. The style of the video is straightforward and documentary-like, focusing on the greenhouse and its contents without any additional embellishments or distractions. The video likely serves as an educational or promotional tool for the greenhouse or the plants being grown inside."
+    ],
+    "video_ids": [
+        "PEXs9UUgqqg_90_0to139"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Plants? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows rows of plants growing inside large white plastic-covered greenhouses. The plants are visible through the translucent covers, and the ground is covered with soil and mulch, indicating active cultivation. This matches the 'Plants' condition as the primary subject of the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a of plants growing inside a, structures containers structureshouses, The plants are clearly and the transparent walls, and they environment is covered with soil, smallch, which a plant. The align the descriptionObjectants' condition as the video focus of the video is}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c01472881b66eedf9be194753f5caf52124c020d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a6e758d18f39aeb16fa1e03835966a8e59b594c1f51a0ad14489fa322f0176ff
+size 108523
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..5c74ab9841fedadc4e55aba1d28581d057707d32
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a bartender is seen preparing a drink at a bar. The bartender, wearing a brown hat and a vest, is pouring a clear liquid from a bottle into a glass. The bar is well-stocked with various bottles of alcohol and glasses. The bartender is smiling and appears to be in conversation with a customer. The customer, wearing glasses, is seated at the bar and is also smiling. The scene is set in a cozy bar with a wooden interior. The bartender's actions suggest that he is in the process of making a cocktail. The overall atmosphere of the video is relaxed and friendly."
+    ],
+    "video_ids": [
+        "mEIxZA_VoVA_2_0to175"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bartender and a customer.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men at a bar: one dressed as a bartender (wearing a vest and bow tie, actively mixing a drink) and another seated as a customer (wearing glasses, watching the bartender). This matches the 'Object(s)' condition of a bartender and a customer.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts a individuals in a bar setting one appears as a bartender,wearing a hat and hatler) holding mixing a drink), and the who at a customer (wearing a and observing the bartender). The setup the descriptionObject(s)' condition of having bartender and a customer in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..265bc09e312c07628d23280103565453e67eb4ee
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:878fd6b959b71aea1c32e6eec1563ec321395ac848dba7b2114cc7fa739ff266
+size 144916
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..6ee6c2dd46b4815c48a71f4c373f70478a951d08
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a black luxury car parked on a street. The car is sleek and shiny, with a large grill and distinctive headlights. The car is parked next to a black fence, and there are other cars parked in the background. The car is stationary, and there are no people visible in the video. The style of the video is a straightforward, unedited shot of the car, with no additional effects or filters applied. The focus is solely on the car, and there is no additional context or background provided. The video is likely intended to showcase the car's design and features."
+    ],
+    "video_ids": [
+        "6R9T_dFElls_2_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black luxury car with a sleek, shiny appearance, large grill, and distinctive headlights.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black luxury car with a sleek, shiny appearance, featuring a large chrome grille and distinctive headlights, which aligns with the description. The car's design and visual details match the specified criteria.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a black luxury car with a sleek, shiny appearance, a a large grill grille and distinctive headlights. which aligns with the description provided The car's design and details elements, the characteristics characteristics.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4bbf27d736a274511649dd3ae67ed363edf928c7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5bbf77bd4b0e7b4d49d7c9bd2aa47de24d1f76943a1a902a45fe6a8b19748a52
+size 250665
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..3bbfaeffadcaa69b4196575b49c008112a99f815
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a motorcycle handlebar and dashboard. The handlebar is black and silver, with a red button on the left side. The dashboard is digital and displays various information. The motorcycle is parked on a gravel road with grass and rocks in the background. The style of the video is a straightforward, clear shot of the motorcycle's controls and display, with no additional action or movement."
+    ],
+    "video_ids": [
+        "WppunsAeEKg_25_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Motorcycle handlebar, motorcycle dashboard, red button on the handlebar? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the motorcycle handlebar, dashboard, and a red button on the handlebar, which aligns with the specified conditions. The presence of additional elements like the mirrors and fuel tank does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a motorcycle handlebar, the, and a red button on the handlebar. which aligns with the ' objects. The presence of additional elements like the background and the tank does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..849892beff6e770ffad2f476b788a1532f492cf6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77ee190bc1c7aec7803356f53c9b9797be1148f7da057ba546be728866263394
+size 218580
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..f58bd2b709e1cc8dd7eab943b51c09e3399bfca2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a vintage fighter plane flying through a clear blue sky. The plane is painted in a camouflage pattern with a star emblem on its side. The propeller is in motion, indicating the plane is in flight. The plane is captured from a side angle, allowing a view of its wings and tail. The overall style of the video is realistic, with a focus on the aircraft and its movement through the sky. The clear blue sky provides a stark contrast to the plane, highlighting its details and design. The video does not contain any other objects or people, focusing solely on the plane and its flight."
+    ],
+    "video_ids": [
+        "Tvb9dDUbBjo_19_0to144"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A vintage fighter plane.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a vintage fighter plane with distinct WWII-era design features, including camouflage paint, a star insignia, and a propeller, all consistent with the description of a vintage fighter plane.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a vintage fighter plane in a features-era characteristics features, such a paint, a prop insignia, and a propeller-driven which of with the description of a vintage fighter plane.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..960a33b8c8f434b6e1e18e8811e70569a41614b1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a8d56b79f1f513218276dae1b99fc0be2f4ca81bda0f5093b7e442b5349f82de
+size 153216
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..25d7c5cd1834763f047713de4f8fa8ad811fd135
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene inside what appears to be an aquarium or zoo enclosure. A polar bear is the central focus, positioned near the glass barrier of its habitat. The bear is partially submerged in clear blue water, with its head and upper body visible above the surface. Its fur is a pristine white, contrasting beautifully with the deep blue of the water. The bear's expression is calm and contemplative, suggesting it might be resting or observing something outside its enclosure.  The background reveals a well-maintained environment designed to mimic the polar bear's natural habitat. There are rocky formations and a sandy area, providing a diverse landscape for the animal to explore. The enclosure is spacious, allowing the bear ample room to move around. Overhead, a large canopy structure provides shade and protection from direct sunlight, ensuring the bear's comfort.  As the video progresses, there are no significant changes in the bear's position or behavior. It remains stationary, occasionally shifting its gaze slightly but maintaining a relaxed demeanor."
+    ],
+    "video_ids": [
+        "c80792c44fec3b08220c45a70397cfe266bce29fa31803ebb073893f362e6cdf"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A polar bear? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a polar bear in an enclosure, partially submerged in water, interacting with a glass barrier. The bear's physical characteristics and behavior are consistent with those of a polar bear, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a polar bear swimming a aquatic, which submerged in water. which with its small barrier. The bear's white characteristics, the are consistent with the of a polar bear, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8a4de08d063fe76dfd04e8f25fd28af7ed4d41dc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b3a35314083b1bc30bafc7cf4b9d72ef5fbe19ae0d574163a7831b313456ed13
+size 265142
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..7f1f69b0bf38090ae74de7af6e46bd0b440baeec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a motorcyclist riding down a street. The rider is wearing a black leather jacket and a black helmet with a visor. The helmet has a camera attached to it, suggesting that the rider is recording their journey. The rider's hands are on the handlebars, and they are wearing gloves. The motorcycle is red and black, and it is in motion. The street is lined with buildings, and there are trees in the background. The sky is clear and blue, indicating that it is a sunny day. The rider appears to be enjoying their ride, and the camera captures the excitement and freedom of the journey."
+    ],
+    "video_ids": [
+        "zVabI3xP4e8_36_0to113"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A motorcyclist on a red and black motorcycle, wearing a black leather jacket, black helmet with a visor, and gloves. A camera is attached to the helmet.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a motorcyclist on a red and black motorcycle, wearing a black leather jacket, black helmet with a visor, and gloves. A camera is clearly attached to the helmet, matching the description. The scene is consistent with the provided details without any contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a motorcyclist on a red and black motorcycle, wearing a black leather jacket, black helmet with a visor, and gloves. The camera is attached attached to the helmet, as the description provided The presence is set with the provided details, any contradictions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cbeb20b8954dca81807c8a90a202291422953a05
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:acfe04681a8729c94107104b7a771066153cca9b51af0d4055d52a62fcbf88c0
+size 280105
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..912224cc7fbf61d3b429c363811b6ded2e64d4b6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a soccer match. A player in a white jersey, representing Real Madrid, is in the midst of a powerful kick, sending the ball flying towards the goal. The player's body is angled towards the ball, showcasing the intensity of the game. The player's white jersey contrasts with the green jersey of the opposing team, adding to the visual appeal of the scene. The background is filled with spectators, their faces a blur of anticipation and excitement. The focus of the video is clearly on the player and the ball, capturing the essence of the sport in a single, powerful frame."
+    ],
+    "video_ids": [
+        "9uRMBb4fwLE_27_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a white Real Madrid jersey and the ball.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player wearing a white Real Madrid jersey, actively engaged in kicking a soccer ball. The core description is accurately represented, with no conflicting elements that contradict the presence of the player and the ball.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a white jersey Madrid jersey, which engaged in a a soccer ball. The player elements of met represented, with the significant elements. would the ' of the player in the ball.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ba9451878bad51e796a9071d0170815c817adfc4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1293-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:610a29d0c4e579038577a4e1a8f63a9a8e59f4c8a6cdf96ba021a611a92d6309
+size 261900
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..9c8451430340013165fb55b2885ad262076f5e12
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up of a monkey's face, showcasing its red fur and expressive eyes. The monkey's fur appears to be wet, possibly from rain or a nearby water source. The monkey's gaze is directed off to the side, giving the impression that it is observing something in its peripheral vision. The background is blurred, drawing focus to the monkey's face. The style of the video is a close-up, slow-motion shot, emphasizing the details of the monkey's fur and facial features. The overall mood of the video is calm and serene, with the monkey appearing to be at ease in its natural habitat."
+    ],
+    "video_ids": [
+        "hdOColvliL0_100_78to207"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A monkey's face? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a close-up of a monkey's face with red skin, which matches the 'Object(s)' condition. Although the video later splits into a two-panel view with an additional image of a person, the monkey's face remains clearly visible and central to the initial and continued focus of the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a close-up of a monkey's face, detailed fur and white align the descriptionObject(s)' condition. The the video is transitions into two split-frame view, a additional monkey of a monkey, the primary's face is the visible and is in the content frames most focus of the video,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..92782c56d1d567e43583313513feb3ad9f9a5079
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:380a359b93e7f0f08c830a446321d887b395c17f742c5118f7514bcc10aaf732
+size 304688
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..692221b1065a2b49a1d278ace9f30cf737b3f6f9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a short, colorful, and vibrant piece featuring a statue of the Virgin Mary surrounded by flowers. The statue is adorned with a pink and gold dress, and it is placed against a backdrop of a stone wall. The flowers surrounding the statue are a mix of red and white, adding a pop of color to the scene. The video is likely a religious or spiritual piece, given the subject matter. The style of the video is simple and straightforward, focusing on the statue and the flowers without any additional elements or distractions. The colors are bright and vivid, making the statue and the flowers stand out against the stone wall. The overall mood of the video is serene and peaceful, reflecting the reverence and respect often associated with religious iconography."
+    ],
+    "video_ids": [
+        "Vj7JQ2frqyA_19_0to154"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A statue of the Virgin Mary, flowers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a statue of the Virgin Mary dressed in purple and gold robes, surrounded by vibrant red and white poinsettia flowers, which matches the 'Object(s)' condition. The presence of a stone wall and text overlays does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a statue of the Virgin Mary, in traditional and gold,, surrounded by a flowers and white flowersoinsettia flowers. which align the descriptionObject(s)' condition of The presence of the statue-like in additionalural does not contradict the core description but}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..eef5351926ba01f839ed74ac67247863795b4337
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:22dbaca36bea408a0a00a10e08e9290a0a64149b4b4b4fd262c170c9fba51892
+size 158685
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..0ec3b04f05c788713c2fb54a752496b1947fc013
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a chef in a professional kitchen, preparing a dish. The chef is wearing a white chef's coat and a black hat, and is focused on the task at hand. The chef is pouring a liquid from a silver pot into a silver bowl. The kitchen is well-equipped, with various kitchen utensils and ingredients visible in the background. The lighting is bright, highlighting the chef's actions and the cleanliness of the kitchen. The overall style of the video is professional and informative, likely aimed at showcasing the chef's skills and the culinary process."
+    ],
+    "video_ids": [
+        "xo0mHkVnVhY_8_21to170"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chef, silver pot, silver bowl? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a chef wearing a white chef's coat and black hat, pouring a liquid (egg mixture) from a silver pot into a silver bowl. The core objects described \u2014 chef, silver pot, and silver bowl \u2014 are all present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a chef in a white uniform's coat and a hat, pouring liquid liquid fromlikely whites) from a silver pot into a silver bowl. The presence elements ( in the, silver pot, and silver bowl \u2014 are all present and accurately depicted in the video.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b1bf06b4d5daf103592d04277d8e07019dedcae7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd73458430973922ed73d353b76b3ea1cc907a33145f984ee56b50391bfa6b62
+size 206660
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..cac6c68c92edecd1c979f532df7f870f86fec8aa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a soccer match. The main focus is a soccer player in a gray and black striped jersey, who is in the midst of a powerful kick. The ball is in motion, heading towards the goal. The player's body is angled towards the goal, indicating the direction of the kick.  In the background, other players are visible, some in red and white striped jerseys, others in blue and white striped jerseys. They are all in various positions on the field, ready to react to the kick. The field itself is a vibrant green, contrasting with the players' colorful jerseys.  The crowd in the stands is blurred, indicating that the focus is on the action on the field. The atmosphere is intense, with the spectators' attention riveted on the unfolding match. The overall style of the video is action-packed and energetic, capturing the excitement and unpredictability of soccer."
+    ],
+    "video_ids": [
+        "jXPacfXZuc8_4_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A soccer player in a gray and black striped jersey, a ball, and other players in red and white or blue and white striped jerseys.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player in a gray and black striped jersey (with 'Fly Emirates' visible), and the player is actively engaged with a ball. Other players are visible in the background wearing red and white or blue and white striped jerseys, matching the description. The scene is consistent with a soccer match, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a soccer player in a gray and black striped jersey,likely the1 Emirates' written on a a player is interacting engaged with a soccer. There players in visible in red background, red and white or blue and white striped jerseys, which the description. The scene is set with a soccer match setting and the additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0d9427cb9c148e6680880aee91144caf575d3d20
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e844ae0c6346e3b19fbe312dc3df0fe3f0302a370427ae743cd49ac93fbe9ba
+size 336586
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..bd5e33f03f4e59dcfbd5617fd5c25cac3e8177d2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young woman with long, wavy blonde hair and a hint of pink at the ends. She is wearing a silver necklace and a sleeveless top. The woman is seated in a stadium-style seating area, which is filled with rows of white seats. The lighting in the scene is soft and natural, suggesting that the video was taken during the day. The woman appears to be deep in thought, as she gazes off to the side. The overall style of the video is casual and candid, capturing a quiet moment in the woman's day."
+    ],
+    "video_ids": [
+        "z67doq_XK9U_84_29to242"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman with long, wavy blonde hair and a hint of pink at the ends. She is wearing a silver necklace and a sleeveless top.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman with long, wavy blonde hair that has a hint of pink at the ends, wearing a silver necklace and a sleeveless top, which matches the description. The background of bleachers does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a young woman with long, wavy blonde hair that has a hint of pink at the ends. which a silver necklace and a sleeveless top. which matches the description provided The presence appears theachers is not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..679109cf941e5c777646ac742fb7c1e6682bc881
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41c647c004ff774e7a2418be1eb999ae985c51a8508688011a083302efe91640
+size 211696
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..0d0f01777c23026b643beb1e62a3db1fa7654d20
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures an exciting moment at a rock crawling event. An orange off-road vehicle, equipped with large tires and a roll cage, is seen navigating a challenging course of large rocks. The vehicle is in motion, with its front wheels lifted off the ground as it climbs over a particularly large rock. The driver, wearing a red shirt, is focused on the task at hand. Spectators, some wearing hats and backpacks, are scattered around the course, watching the vehicle's progress with interest. The event is taking place in a natural setting, with trees and rocks forming the backdrop. The video is shot from a side angle, providing a clear view of the vehicle's maneuver and the spectators' reactions. The overall style of the video is dynamic and action-packed, capturing the thrill and excitement of the event."
+    ],
+    "video_ids": [
+        "dh9QWPuxZoE_127_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An orange off-road vehicle, the driver in a red shirt, and spectators.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows an orange off-road vehicle navigating a rocky terrain, with a driver visible in a red shirt. Spectators are also present around the vehicle, observing the event. All core elements described in the condition are accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts an orange off-road vehicle navigating a rocky terrain. which a driver visible in a red shirt. Thereators are also present, the vehicle, watching the action. The elements elements of in the condition are present represented in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9f9304b7fb083fa318cb0d76a48d675ee8058d47
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:000f4a1bb6933a000834cacfbdd52d25567751aa441a00b974c20d25fc819f88
+size 222653
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..e67e6cd62ae75707f6c578f466713bc010a514fb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a scene in a parking lot where a silver BMW car is parked next to a black SUV. The car is parked in a spot that is adjacent to a building. The car is facing the camera, and the SUV is parked behind it. The car is shiny and well-maintained, indicating that it is likely a newer model. The SUV, on the other hand, appears to be older and has a more rugged design. The building in the background is large and has a flat roof, suggesting that it might be a commercial or industrial building. The parking lot itself is paved and appears to be well-maintained. The overall style of the video is straightforward and documentary-like, with no additional embellishments or effects. The focus is solely on the cars and their immediate surroundings."
+    ],
+    "video_ids": [
+        "uwfULmnCwnQ_7_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A silver BMW car and a black SUV.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a silver BMW sedan and a black SUV parked next to each other in a lot, which matches the core description. Additional cars and background elements do not contradict this primary focus.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a silver BMW car and a black SUV, in to each other in a parking. which align the description description of The elements and a elements are not contradict the main focus.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3302663ce7ffaabc34e8eaa8548ae41122dcd0c5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:452d37756aa975b2aa53deb742e2a9442fef5d4a25556d54d9ec848dec19fe13
+size 169999
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..c4b71afafd5e8c7e691ceffad4fa6023b58694e7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse shot of a city skyline, showcasing the progression of the sky from day to night. The skyline is composed of numerous tall buildings, each varying in height and architectural design. The buildings are clustered closely together, creating a dense urban landscape. The sky transitions from a clear blue during the day to a deep, starry night. The video captures the essence of a bustling city, with the towering structures standing tall against the changing sky. The overall style of the video is a blend of urban architecture and natural elements, creating a harmonious balance between the man-made structures and the natural world."
+    ],
+    "video_ids": [
+        "APH2HpyjlP4_0_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Numerous tall buildings? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features numerous tall buildings, viewed from a low-angle perspective looking upward. The buildings are clearly visible and dominate the frame, fulfilling the 'Numerous tall buildings' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features numerous tall buildings, which against a low angle perspective that up. The buildings are illuminated visible and dominate the lower, fulfilling the 'Objectous tall buildings' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e89ee5b6c4a3a48e23d34a3d77bb5a4943b5783f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1294-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0952b94b75f3e9efcc1e7eafacb4d6282da6e3624baefc7f1e4582f98cdb6c5
+size 272632
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..ddf71c19dfcced492737055fffe12b29f9ef81cd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are in a kitchen, preparing a cake. The woman is wearing a pink apron and is standing behind the counter, while the man is standing beside her. They are surrounded by various kitchen items such as bowls, spoons, and cups. The kitchen is well-equipped with an oven and a refrigerator. The counter is filled with ingredients and utensils, indicating that they are in the process of baking a cake. The man and woman seem to be working together, suggesting a collaborative effort in the cake-making process. The overall atmosphere of the video is warm and inviting, with the focus being on the cake-making process."
+    ],
+    "video_ids": [
+        "CdcjB4cp4Z4_29_0to189"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman standing in a kitchen setting, engaging in what appears to be a cooking or baking demonstration. Both individuals are visible and central to the scene, fulfilling the 'Object(s)' condition as described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a man and a woman in in a kitchen,. engaging in an appears to be a cooking activity baking activity. The individuals are positioned and are to the scene, fulfilling the 'Object(s)' condition as described.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dd4f1b9b5f152d5e0ec9c0a975867706a487d96a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:82dfc1aa5aee2b3aab55afcffc6e69ad7ff880823b3a8a64694a7997f8a547b3
+size 126832
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..286c503ee9f7569dad536a340f23bf595d12608b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse of a table setting, showcasing a collection of objects. The table is adorned with a black and white patterned tablecloth. The objects include a large crystal, a stack of tarot cards, and several candles. The crystal is placed on the left side of the table, while the tarot cards are stacked on the right side. The candles are scattered across the table, with some placed near the crystal and others near the tarot cards. The objects are arranged in a symmetrical manner, creating a balanced and harmonious composition. The lighting in the video is soft and warm, highlighting the textures and colors of the objects. The overall style of the video is serene and mystical, evoking a sense of tranquility and introspection."
+    ],
+    "video_ids": [
+        "H2gTEgHlO24_8_40to203"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large crystal, a stack of tarot cards, and several candles.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a large crystal (a crystal ball on the left), a stack of tarot cards (in the center-right with a crystal cluster on top), and several lit candles (around the setup). These core elements are prominently featured and accurately match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a large crystal,a clear pyramid), the left), a stack of tarot cards (on the foreground),), a visible ball on top), and several candles candles (around the crystal). The elements elements match present displayed, match represent the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..16bdc62535b8b5d1fef7d5f948b88743c9f370e2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf4d7dbe727a47a9605f2165f292d1399e84c74b311d6819844ed207964a7c66
+size 175365
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..9525c9fc8b10ecfd3633cc52e22d205126e7df8e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman standing in a colorful room filled with various swimming pool toys. She is wearing a red shirt and has a smile on her face. The room is vibrant with a variety of colors, including red, blue, and green. The toys are scattered around the room, adding to the playful atmosphere. The woman appears to be enjoying herself in this lively environment. The video captures a moment of joy and fun, set against the backdrop of a bright and cheerful room."
+    ],
+    "video_ids": [
+        "EE6M_c344Gs_0_0to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, swimming pool toys? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman standing in what appears to be a swimming pool area, with various swimming pool toys visible in the background, such as colorful foam noodles and other pool equipment. These elements align with the 'Object(s)' condition specified.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a woman in in a appears to be a room pool setting, holding various colorful pool toys scattered in the background. such as colorful floats balls and balls pool floats. The elements fulfill with the 'Object(s)' condition of in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..69d2a778d537ef81a900bff908416ac713fcd2cc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f470f7c9f68c0464a28a5f5206090bcfb493abb92ed92b1a1f4aecef9d04e49e
+size 117390
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..e1877e756fae87960b309ef1136579cfeb5e74f3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a professional ice hockey player in action. The player, dressed in a blue jersey with the number 8, is seen holding a hockey stick and skating across the ice. The player's movements are fluid and dynamic, showcasing the agility and speed required in the sport. The ice beneath the player's skates glistens under the bright lights of the arena, reflecting the intensity of the game. The player's focus is evident, eyes locked on the goal, ready to strike. The background reveals a bustling hockey rink, complete with advertisements and spectators, adding to the atmosphere of a high-stakes match. The video is a dynamic snapshot of a moment in the game, capturing the player's skill and the excitement of the sport."
+    ],
+    "video_ids": [
+        "a6jpJMSilKg_53_0to155"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A professional ice hockey player in a blue jersey with number 8, holding a hockey stick.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a professional ice hockey player wearing a blue jersey with the number 8 visible on the sleeve, holding a hockey stick. The player's attire, equipment, and setting are consistent with the description, even though some background elements (like advertisements) are present, which do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a professional ice hockey player wearing a blue jersey with the number 8, on the back. holding a hockey stick. The player is attire and including, and the ( consistent with the description of indicating though the details elements likelike the and are not, they are not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5b293d8176b5dda83891d74ea04467a40757476d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aa4fcee094b4bba3597aa508e0b2778802fcd5b3a1ea61f1251d0de176690e3a
+size 297990
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..c2f2802b3a2aeb76fccd6cfacd33a1d1a2f5edee
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of triumph and joy on a football field. A player, wearing a blue jersey with the number 44, is the focal point of the video. He is seen celebrating a victory, his arms raised high in the air, a wide smile on his face. His teammates, also in blue jerseys, are seen in the background, sharing in the celebration. The stadium is filled with spectators, their faces a blur of excitement and anticipation. The video is a dynamic snapshot of a moment of victory, capturing the energy and emotion of the game."
+    ],
+    "video_ids": [
+        "nEJEjSsEExk_5_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a blue jersey with number 44, his teammates in blue jerseys, and stadium spectators.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a blue jersey with the number 44, consistent with the Buffalo Bills team. Another teammate in a blue jersey is visible behind him. The background clearly depicts a crowded stadium with spectators, fulfilling all the specified conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a blue jersey with the number 44, which with the description Bills' colors The player in a blue jersey is partially in him, The background includes shows a stadium stadium with spectators, fulfilling the the conditions conditions.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..819f5007ffc5ce0f8b2caf6cfe7aa14609cc7116
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7bd8edc5db43bd821031f59bf21a72b3ce40359c6ceb751e4368e0b51b4ea49a
+size 176658
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..71600bd7f7b413b9fd98ab6e113f0ab445379bbb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with long blonde hair, wearing a red and black plaid shirt, and a black choker. She is speaking and looking to the side. In the background, there is a blurred image of a vase with pink and yellow flowers. The style of the video is casual and informal, with a focus on the woman's expression and the soft, blurred background. The lighting is soft and natural, suggesting an indoor setting. The overall mood of the video is relaxed and conversational."
+    ],
+    "video_ids": [
+        "YQN-x7s6vz4_10_0to160"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Woman with long blonde hair, wearing a red and black plaid shirt, and a black choker.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with long blonde hair, wearing a red and black plaid shirt, and a black choker. These elements match the description provided in the 'Object(s)' condition. The background elements, such as flowers and a window, do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with long blonde hair, wearing a red and black plaid shirt, and a black choker. The elements match the description provided, the questionObject(s)' condition. The background,, such as the, a vase, do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c3f03aeb50edad5cc477778b8cf3deacfe3f2b00
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8e6877eec3dabc1febf388e4f4538c5ebec36caabe02df3d66b100b566e8abed
+size 118815
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..4052a8b1cccda8f32a893e5e4706cdf95b840ee5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and glasses, wearing a yellow sweater and a blue shirt underneath. He is seated in front of a blue background with a lighted border. The man appears to be engaged in a conversation or an interview, as suggested by his attentive gaze and the presence of a microphone in front of him. The overall style of the video is professional and polished, with a focus on the man and his interaction with the interviewer or the audience. The lighting is bright and even, highlighting the man's features and the microphone without creating harsh shadows. The blue background provides a calm and neutral backdrop that does not distract from the man or the conversation."
+    ],
+    "video_ids": [
+        "6Z5fTYfVpaQ_31_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man with a beard and glasses, wearing a yellow sweater and a blue shirt, microphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and glasses, wearing a yellow sweater over a blue shirt, and a microphone is visible clipped to his shirt. These elements match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and glasses, wearing a yellow sweater over a blue shirt. sitting there microphone is present in to his sweater. The elements match the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d71b4046006b89cdc6c07ec2b8d75ecd0a8c0073
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f31e846422d655a9d9294bc388a1b1d2a0741d8f4813f96f6daafafce85c1817
+size 127720
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..8e72247c9af9642bb10294cebe64f6e1583f3eb9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman with blonde hair is seen singing passionately into a microphone. She is dressed in a black sequined dress and her hands are raised in a gesture that suggests she is deeply engaged in her performance. The background is a blur of light, creating a dramatic effect that highlights her as the focal point of the scene. The video captures the intensity and emotion of her performance, making it a captivating visual experience."
+    ],
+    "video_ids": [
+        "MOgwB0Ks8zc_8_0to130"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with blonde hair, a black sequined dress, and a microphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair, wearing a black sequined dress, and singing into a microphone. These elements match the core description provided. The background and additional elements do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair, wearing a black sequined dress, and holding into a microphone. The elements match the description description provided. The presence and lighting lighting, not contradict the main.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..45e73608b85df3b3862bb092dafe7ec5724984df
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1295-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef8757d575ddc5087051650c497fae3a341b48d266f9c5ebe1d5e52a868c17bf
+size 208645
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..bb879cff1778455424d8305b95aa72943a575cbb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two women are standing in front of a red pickup truck on a dirt road. The woman on the left is wearing a green shirt and blue jeans, while the woman on the right is wearing a red shirt and blue jeans. They are both smiling and appear to be in a good mood. The truck is parked on the side of the road, and there is a fence in the background. The sky is overcast, and the ground is covered in grass. The overall style of the video is casual and friendly, with a focus on the interaction between the two women and their surroundings."
+    ],
+    "video_ids": [
+        "EO273WqQ2Qo_11_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two women, one in a green shirt and blue jeans, the other in a red shirt and blue jeans.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two women standing in front of a red truck on a dirt road. The woman on the left is wearing a green shirt and blue jeans, and the woman on the right is wearing a red patterned shirt and blue jeans. This matches the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two women standing in front of a red truck. a cloudy path. One woman on the left is wearing a green shirt and blue jeans, while the woman on the right is wearing a red shirted shirt and blue jeans. The matches the description provided, the questionObject(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a28134ddfe59fc43c3edda2866ae8da225e4fd34
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a087bc573a8f75cf0b18eb4279879c2cb3b66be4511177abb8b74c238336630
+size 197706
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..81b9c3a88bfac618ea1c604c6f243a51f1c6004b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a vlog style, featuring a small, grey dog with floppy ears and a black nose. The dog is on a leash, standing on a gravel surface. In the background, there is a parked car and a trailer. The dog appears to be attentive and curious, looking off to the side. The overall atmosphere of the video is casual and relaxed, with a focus on the dog's interaction with its environment."
+    ],
+    "video_ids": [
+        "C07BuaheUtw_6_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small, grey dog with floppy ears and a black nose. The dog is on a leash and appears attentive and curious.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small grey dog with floppy ears and a black nose, wearing a leash. The dog is attentive and appears curious, looking around and towards the camera. The background elements (vehicles, RVs) do not contradict the core description of the dog.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small, dog with floppy ears and a black nose, which a red. The dog appears standing and curious curious, as around and moving the camera. The presence includes,a and scattered,, do not contradict the core description of the dog.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ff30dc54bd9db5a2b83603ed97c672eabf7ff111
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ccd3a9ea9e24fdb64a733d1b96fca0c1385ea254890e4a88393b47fee75a044e
+size 201021
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..a52817036b09721ebdce696ed228c872be86a8fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a red motorcycle parked on a street. The motorcycle is sleek and shiny, with a black seat and handlebars. The front of the motorcycle is prominently displayed, showing the headlight and the front wheel. The motorcycle is parked next to a blue car, which is also visible in the frame. The background of the video shows a brick building and a clear sky. The style of the video is realistic and it captures the details of the motorcycle and its surroundings."
+    ],
+    "video_ids": [
+        "2nJOo6T_5nk_65_0to134"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red motorcycle and a blue car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red motorcycle (specifically a Ducati Panigale V4 R) and a blue car parked nearby, fulfilling the 'Object(s)' condition. Additional background elements like buildings and utility poles do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a red motorcycle inwithally a redati)igale V4)) and a blue car in in. which the 'Object(s)' condition. The elements elements like the and other poles are not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..78aaa25c7f81ae0bab48d8386f9c8eb9e03c87fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:beb4399d2939cc147dc6325b8753953c3c99d03d5d7565aba9530eebb4a48513
+size 200200
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..58fc38f5c7c7d1c3d962b0394372a1ce97ccbd8d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a vibrant nighttime scene illuminated by colorful lights. The setting appears to be an outdoor area, possibly a park or garden, adorned with various light installations. The sky is dark, indicating it is nighttime, and the moon is prominently displayed in the upper center of the frame, casting a soft glow over the scene.  In the foreground, there is a body of water reflecting the lights, adding a serene element to the otherwise lively display. The water's surface is dotted with small, twinkling lights, creating a magical atmosphere. Surrounding the water, there are numerous trees and plants, each decorated with lights in shades of green, blue, and white. These lights give the impression of a forest or garden at night, with the trees standing tall and illuminated.  Scattered throughout the scene are various animal figures, including giraffes and elephants, also lit up with colorful lights. These figures add a whimsical touch to the overall ambiance. The giraffes are positioned near the right side"
+    ],
+    "video_ids": [
+        "9c11f880ca2064bceea20afaff3d06346669a15da9eecb1309b9f327d8ce2d01"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Body of water, trees/plants, giraffes, elephants? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a body of water (reflecting the lights), trees/plants (green lighted trees), giraffes (yellow and orange lighted giraffes), and elephants (blue lighted elephant on the left). All specified objects are present and identifiable despite the stylized, AI-generated appearance.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a body of water,reflecting the colorful and trees/plants (ill anded trees), andaffes (sil light orange lighted figuresaffes), and elephants (pink anded elephants). the right). The these objects are present and do in the artisticized, colorful-generated nature.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..41062c4c107e921022e4b51d0acab75d93287849
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd5e0f342050306f1d566521d86ed6dc1774ec928b710da13d7f38e55c23f50f
+size 192292
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..df60c66b6e11cc0992d47807f789ab44017d89d0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young man with blonde hair, wearing a grey t-shirt with the word \"SHOREHAWK\" printed on it. He is seated in a room with a white wall and a window in the background. The man appears to be speaking, as his mouth is open and he is looking directly at the camera. The lighting in the room is soft and natural, suggesting it might be daytime. The overall style of the video is casual and candid, capturing a moment in the man's life."
+    ],
+    "video_ids": [
+        "3VANVBUX-Uk_32_0to144"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man with blonde hair, wearing a grey t-shirt with the word 'SHOREHAWK' printed on it. He is seated and speaking, with his mouth open and looking directly at the camera.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man with blonde hair, wearing a grey t-shirt with 'SHOREHAWK' printed on it. He is seated and appears to be speaking, with his mouth open and looking directly at the camera. These elements match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man with blonde hair, wearing a grey t-shirt with theSHOREHAWK' printed on it. He is seated and appears to be speaking, with his mouth open and looking directly at the camera. The elements match the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b10217527d0a85229d5a17ed5473066e7e313c0c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6e34361310a542d1a09536a4b27f4454e78b856e54c842c5e5a3d5358e8ac62c
+size 154386
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..167763a089eefaec802087e2cfdf4c10cd7f6e55
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view inside an air fryer, showcasing a batch of small, square-shaped food items being cooked. These items appear to be breaded or coated, possibly chicken nuggets or similar fried snacks, arranged neatly on the air fryer's tray. The air fryer's glass door is closed, allowing a clear view of the contents inside. The background reveals a kitchen setting with tiled flooring, a dining table with chairs, and various kitchen items such as a bottle and a box placed on the table. The lighting is consistent throughout the video, highlighting the golden-brown color of the food items as they cook. There are no visible characters or significant changes in the scene; the focus remains on the cooking process within the air fryer."
+    ],
+    "video_ids": [
+        "05b49896dcec83067c363cba95e49664c4ed42801fe04aef88161904979b91db"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A batch of small, square-shaped food items (possibly breaded chicken nuggets) arranged on the air fryer's tray.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a batch of small, square-shaped food items arranged on the air fryer's tray, which matches the description. The items appear to be breaded and are evenly distributed across the tray, consistent with the idea of chicken nuggets or similar food items. The surrounding kitchen environment does not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a batch of small, square-shaped food items that on the air fryer's tray. which matches the description of The food appear to be breaded, are evenly spaced across the tray, consistent with the typical of chicken nuggets or a food items being The setting environment setting and not contradict the core.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3d48341cb627c3c219d505ad80b3bf67a4f4bbf7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ca5d73a4dbb51a3a756d44811ea9663964568a0980e03e9c52f32ac8d324d04b
+size 74054
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..ca84b5ac1ba150417dc3480c5d908cb99f3dbbf0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is an aerial view of a luxurious estate, showcasing its various amenities and the surrounding landscape. The estate features a large, well-maintained tennis court, a swimming pool with a curved shape, and a spacious lawn. The architecture of the main house is modern and elegant, with multiple levels and a mix of flat and sloped roofs. The property is surrounded by lush greenery, including palm trees and other tropical plants, which contribute to the overall serene and upscale atmosphere. The aerial perspective provides a comprehensive view of the estate's layout and design, highlighting the thoughtful integration of the man-made structures with the natural environment."
+    ],
+    "video_ids": [
+        "5a3v7vCCUdM_4_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Tennis court, swimming pool, lawn, main house? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a tennis court, a swimming pool, a well-maintained lawn, and the main house, all of which are prominently featured in the aerial view. These elements are consistent with the requested 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a tennis court, a swimming pool, a well-maintained lawn, and a main house. which of which are consistent featured in the scene view of The elements match consistent with the ' objectsObject(s)' and.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8589b7a21f84296ff315089a68371d867b2e6803
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f485fbfd98eaff29c107a4e2f2b67a9a508071b2083f5e486cfc9e3aad721ac
+size 250292
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..67f7e8153fb52c05c73d1c3238939001b49dde2b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a thrilling journey through the desert in a green jeep. The driver, dressed in a beige shirt, is seen holding the steering wheel firmly, navigating the vehicle through the sandy terrain. The jeep is equipped with a spare tire mounted on the back, ready for any unexpected situations. The dashboard of the jeep is visible, indicating the vehicle's speed and direction. The vast expanse of the desert stretches out in the background, with sand dunes visible in the distance. The clear blue sky above adds to the serene yet adventurous atmosphere of the scene. The video is a testament to the thrill of off-roading and the beauty of the desert landscape."
+    ],
+    "video_ids": [
+        "DYXrh7mcKtY_67_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Green jeep, spare tire, driver (dressed in a beige shirt), dashboard? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a green jeep with a spare tire mounted on the back, a driver wearing a beige shirt, and a visible dashboard with gauges. All elements described in the condition are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a green jeep driving a spare tire mounted on the back, a driver wearing a beige shirt, and a dashboard dashboard. auges and The elements in in the condition are present in match with the video content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cf0cbff00ca6b55dfacdd8491587b6d30713118a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1296-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2303997975d4e77da94575ec95bfc9d8b5306bc79eb931ef1e4afd7d373c87c1
+size 147920
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..4fdc3924340a5c7116e4443f933cf682141d33fb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a segment from a daytime talk show. The host, a woman in a pink sweater, stands in the center of the stage, engaging with the audience. She is surrounded by a diverse group of people, some of whom are seated on a couch, while others are standing or sitting on chairs. The audience members are actively participating in the conversation, with some smiling and nodding, while others are actively speaking. The set is designed with a blue and yellow color scheme, and the logo of the network is visible in the corner. The overall style of the video is casual and conversational, with a focus on the interaction between the host and the audience."
+    ],
+    "video_ids": [
+        "4_AJqbAapBU_18_122to334"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Host (woman in a pink sweater), audience members (some seated on a couch, others standing or sitting on chairs)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a pink sweater standing in front of an audience, which includes people seated on couches and chairs. The setup matches the description of a host and audience members as specified in the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a woman in a pink sweater standing and front of a audience, which align people seated on aes and chairs. This setup align the description of a host addressing audience members, described. the question.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..27f4b7f1a96d6eaad8d9d17a1ddbc6b11fb3cef0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f62d31fd3cbc0410a0e48d3a98d133b96abb84aa2ecade89c700d204884886a5
+size 128835
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..496789eb4bef21ac98f8a82fc26c45fd3c9a49a1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is an aerial view of a city, showcasing its urban landscape and infrastructure. The style of the video is a bird's eye view, providing a comprehensive perspective of the city's layout and design. The city is densely populated with buildings of various sizes and shapes, indicating a mix of residential, commercial, and industrial areas. The roads and highways are clearly visible, connecting different parts of the city. The video captures the city's growth and development, highlighting the urban planning and architecture. The colors in the video are natural, reflecting the city's environment and the materials used in the construction of the buildings. The video is a visual representation of the city's evolution and its impact on the urban landscape."
+    ],
+    "video_ids": [
+        "XKjJTsQunxY_24_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Buildings, roads/highways? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows multiple buildings of various sizes and designs, as well as roads and highways with visible traffic. These elements are central to the visual content and align with the 'Object(s)' condition specified.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a buildings of various sizes and designs, which well as a and highways that visible traffic. The elements are consistent to the depiction description and align with the 'Build(s)' condition described in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dc51bb2e4eee777514e4dc65598a9cf90577e673
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:19cfc1fa0ffc19e23650a858b68b8cd0ee1cb6af9810e0a786cc151fb2bb407c
+size 272813
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..8238c2480432be92c07b9912f9461f7db904cbf0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a butterfly perched on a yellow flower. The butterfly, with its wings spread wide, is the main focus of the video. It is seen in three different positions: first, it is seen landing on the flower; second, it is seen resting on the flower; and third, it is seen taking off from the flower. The background of the video is blurred, but it appears to be a garden with green plants. The video is shot in a way that the butterfly and the flower are in sharp focus, while the background is out of focus. The overall style of the video is a close-up nature shot, capturing the beauty of the butterfly and the flower in detail."
+    ],
+    "video_ids": [
+        "-PY9PlRum28_7_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A butterfly and a yellow flower? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a butterfly with orange, black, and white patterns resting on a bright yellow flower. The background contains green foliage and a blurred structure, but these do not contradict the core description of a butterfly and a yellow flower. The scene is consistent with the requested object(s).\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a butterfly per orange and black, and white patterns per on a yellow yellow flower. The background is green foliage, additional blurred effect, which these elements not contradict the core description of the butterfly and a yellow flower. The presence is consistent with the given object condition)\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cc86e17868e4c4e22606961d4d10c2efd350463a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c65f820095f95610c26f9e600ece66bb4697c2fe359701aa6bda3cdd38b847ef
+size 178574
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..7f9fb5ad46e59eb3b26bc57a25e9d1b4b5af2e23
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man named Jay Stubbarg, who is standing in a news studio. He is dressed in a blue shirt and tie, and has a beard. The studio has a blue background with the words \"newsy\" written on it. There is also a TV screen in the background. Jay is speaking to the camera, and his expression suggests that he is making a serious point. The overall style of the video is professional and news-like."
+    ],
+    "video_ids": [
+        "5Zfk67-bulM_1_19to202"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Jay Stubbarg, a man dressed in a blue shirt and tie with a beard. A TV screen in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man who matches the description: he is wearing a blue shirt and tie and has a beard. The background includes what appears to be a TV screen or monitor, consistent with an office or newsroom setting. The name 'Jay Strubberg' is displayed, which closely matches the name 'Jay Stubbarg' in the query (likely a minor spelling variation). There are no significant contradictions with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed is the description of he is dressed a blue shirt and tie, has a beard. The background includes a appears to be a TV screen, a, which with the office or studio studio setting. The presence 'Jay Stubarg' is not on which could matches the name 'Jay Stubbarg' in the description.with a minor typo difference). The are no significant contradictions or the given description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..65693fe87a120d4b6a99607e1f158ab8ee942e70
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4cf0e8e0704448e7e4865937df9e1fc4f7b212ced83e00c907b13beb1cde37f4
+size 145154
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..a257d0ee52e855b42366e5ee54ff4b6fb8a06dd8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a collection of star-shaped decorations hanging from a string against a white wall. The stars are made of different materials, including wood and paper, and they are arranged in a way that creates a sense of depth and dimension. The stars cast shadows on the wall, adding to the visual interest of the scene. The video is likely to be a simple, yet elegant decoration for a room or event. The style of the video is minimalistic and focuses on the beauty of the star decorations and their shadows."
+    ],
+    "video_ids": [
+        "QBXtunxyov0_0_0to116"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Star-shaped decorations? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features star-shaped decorations, including wooden and white star cutouts hanging and casting shadows on a wall. These objects clearly match the description of 'Star-shaped decorations'.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features star-shaped decorations hanging which both and paper star ornamentsouts hanging from casting shadows on a wall. The elements clearly fulfill the description of 'Star-shaped decorations'. The}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..101cc9b55bc3a158ca00f7f34253fc0492905373
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d5f795b5a83bd9341b1671b25a9874ee82763709f084f1c2c2b60ec712d2596
+size 168544
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..ef65e57032edff6ab43c855ad6b092ee46be03a1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated family scene. A man, a woman, and three children are sitting on a pink couch with a floral pattern. The man is wearing a blue plaid shirt and orange pants, while the woman is wearing a pink dress with a white blouse and glasses. The children are dressed in colorful outfits, with the girl in the middle wearing a pink dress with a white blouse and glasses. The room has a pink wall with star stickers and a blue lamp on a table. The family appears to be engaged in a conversation, with the man and woman looking at the children. The overall style of the video is bright and cheerful, with a focus on family bonding and interaction."
+    ],
+    "video_ids": [
+        "75eje3XKa7w_412_0to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a woman, three children, a pink couch, a blue lamp, a table, star stickers, a pink wall? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a man, a woman, three children, a pink couch, a blue lamp, a table, star stickers, and a pink wall. All these elements are present and consistent with the description, with no contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a man, a woman, three children, a pink couch, a blue lamp, a table, star stickers, and a pink wall. The these elements are present and match with the description. and no contradictions or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..03d806d9ca52a7b2d357c4fe0fdf3850ceec7576
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb4e4740214c43cec3733ed3b28d71c08d0836fc8bf55eb569807bafc010947a
+size 104198
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..af8c796a0beb55d7fbbf526ddf2e6ed8906d0247
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene lakeside scene, framed by lush green trees on both sides. The water is calm, reflecting the sky above, which is mostly clear with a few scattered clouds. A prominent feature in the foreground is a textured wooden railing, adding a rustic element to the composition. To the left, a red structure, possibly part of a building or a decorative element, stands out against the natural backdrop. As the video progresses, fish can be seen swimming near the surface of the water, creating gentle ripples that disturb the otherwise tranquil surface. The camera remains stationary throughout, focusing on the peaceful interaction between the still water and the lively aquatic life beneath it. The overall atmosphere is one of quiet beauty and natural harmony."
+    ],
+    "video_ids": [
+        "be76dca05cf69ea03721ca07798f8aad466eb9ffa168fcabac32a90a9e1adacd"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Textured wooden railing, red structure, fish? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a textured wooden railing in the foreground, a red structure (likely part of a bridge or viewing platform) on the left, and fish swimming in the water. These elements are clearly visible and match the described 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features ful a textured wooden railing in the foreground, a red structure onpossibly a of a gaze or dock platform) on the left side and the swimming in the water. The elements align consistent visible and match the description objectsObject(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..52ed89ac45526e6377005e8ed112f682d0a73b26
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fbe8bd4b656fcf97a9958f802fcdd7e73c0aef58e5609f239d8a3e778076036e
+size 247423
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..76ff3b28e4d0be19adbb9622e7093ba6334c7fcd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman in a white and red ski suit, wearing a pair of goggles, standing against a green background. She is holding a helmet in her hand. The style of the video is a promotional advertisement for a product called \"BIOSUITE\". The woman is the main subject of the video, and she is positioned in the center of the frame. The background is simple and uncluttered, allowing the viewer to focus on the woman and the product. The overall tone of the video is clean and professional."
+    ],
+    "video_ids": [
+        "UCd2mYGHw1s_76_39to162"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a white and red ski suit, a pair of goggles, and a helmet.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a white and red ski suit with visible red accents and black lines, along with goggles on her head and holding a helmet. These elements match the description provided, even though the context suggests a futuristic or space-themed outfit rather than traditional ski gear.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a white and red ski suit, a logos accents, a gloves. which with a and her head. a two helmet. The elements match the description provided, indicating though the helmet of the digitally or AI-themed setting, than a skiing gear.}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a2a89f095ee646a389bda8986bfebe7b6aa1cef0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1297-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86ab4a0ab4b6a3a74e9ab39e397472e827488a8f3548643678333290db8f8a33
+size 215201
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..d3d1fc2bdc9e6eadf39d4d24b99f66be7b88f1e7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene hike through a lush, green forest. The camera follows two hikers as they traverse a winding path, surrounded by vibrant foliage. The hikers are dressed in casual hiking attire, and one of them carries a backpack. The forest is dense with trees and shrubs, and the path they follow is well-trodden, suggesting it's a popular hiking route. In the distance, majestic snow-capped mountains rise against the clear blue sky, adding a sense of grandeur to the scene. The overall style of the video is naturalistic, with a focus on the beauty of the landscape and the peacefulness of the hike. The camera work is steady, allowing the viewer to fully appreciate the details of the forest and the mountains. The lighting is bright and natural, enhancing the colors of the foliage and the blue of the sky. The video does not contain any text or music, allowing the viewer to fully immerse themselves in the experience of the hike."
+    ],
+    "video_ids": [
+        "wTjE6G5GTUU_70_66to195"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two hikers dressed in casual hiking attire, one carrying a backpack.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two individuals in what appears to be casual hiking attire, with one visibly carrying a backpack. They are positioned in a natural, mountainous environment, consistent with hiking activity. The description of their attire and gear aligns with the visual evidence, even though the video is AI-generated, the core elements match the requirement.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two individuals walking casual appears to be casual hiking attire, walking one of carrying a backpack. The are walking on a natural setting outdoorous setting, which with the.. The presence of ' attire and the aligns with the typical elements in indicating though the video is AI-generated, it core elements match the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4cc9c50bb20f0870eed85a66c7b6dea0180d05f5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b656c3bfa6597e1839aff716c653441c8877662a5e1c6d5ac5b94bd06cc54d9c
+size 353602
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..535d65428cb2dce7928ff965d05079ce99886092
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a futuristic car on display in a showroom. The car is sleek and aerodynamic, with a shiny silver exterior that reflects the surrounding environment. The car's design features a large rear wing and a distinctive front grille, giving it a high-tech and sporty appearance. The car is positioned on a blue carpet, which contrasts with the silver of the car and highlights its presence in the showroom. The showroom itself is spacious and well-lit, with large windows that allow natural light to flood in and illuminate the car. The overall style of the video is sleek and modern, with a focus on the car's design and the showroom's ambiance."
+    ],
+    "video_ids": [
+        "lQWxfB-9frA_25_0to195"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A sleek, aerodynamic car with a shiny silver exterior, a large rear wing, and a distinctive front grille.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a sleek, aerodynamic car with a shiny silver exterior, a large rear wing, and a distinctive front grille. The car's design matches the description, and no conflicting elements are present that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a sleek, aerodynamic car with a shiny silver exterior, which large rear wing, and a distinctive front grille. The car's design is the description provided and the additional elements are present in would the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1f20099e35bd44b8ab2fa776e1e86fff79408327
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:369edc098654c1eecf0ef1bcdc4ab6d08602800ed7f7c79caaa1c9025b6baae8
+size 230914
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..b0227624fd3730296b20a407f412d34eb11a21c3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene mountain landscape, with a wooden pole standing prominently in the foreground. The pole, which appears to be a marker or post, is surrounded by a small pile of rocks, adding a rustic charm to the scene. The background is dominated by majestic mountains, their peaks reaching towards the clear blue sky. A winding dirt path cuts through the grassy field, inviting viewers to imagine a journey through this picturesque setting. The overall style of the video is naturalistic, with a focus on the beauty of the landscape and the simplicity of the wooden pole and rock pile. The colors are vibrant, with the green of the grass and the blue of the sky providing a striking contrast to the brown of the mountains and the wooden pole. The video is likely to evoke feelings of tranquility and awe in the viewer, as they take in the grandeur of the mountains and the peacefulness of the field."
+    ],
+    "video_ids": [
+        "zpHtMxqoCWk_74_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A wooden pole in the foreground, surrounded by a small pile of rocks; a winding dirt path through the grassy field.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a wooden pole in the foreground, embedded in a small pile of rocks, and a winding dirt path through the grassy field. These elements are consistent with the description, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a wooden pole in the foreground, surrounded in a small pile of rocks. which there winding dirt path through a grassy field. The elements match consistent with the description provided and there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9b06254dee2112647369123201c9062596ab612b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aed9ac84f7a8a4c229b65987defb0828eeb1a20f7c07e977c4da2888a537a063
+size 79226
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..ba23ff79fb1bacd237a4a9c5b9083f8a7cc6936f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man sitting at a table in a restaurant, enjoying a meal. He is wearing a blue polo shirt and is giving a thumbs-up gesture, indicating his satisfaction with the food. The table is set with a white plate of food, a cup, and utensils. The restaurant has a modern and elegant interior, with white walls and green accents. The lighting is soft and ambient, creating a warm and inviting atmosphere. The man appears to be in good spirits, smiling and engaging with the camera. The video captures a pleasant dining experience in a stylish and comfortable setting."
+    ],
+    "video_ids": [
+        "Pxm6UIDvTko_67_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a table, a white plate of food, a cup, and utensils.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man seated at a table with a white plate of food, a cup, and utensils (a fork and knife). The scene is set in a restaurant, and all the specified objects are present and accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man sitting at a table with a white plate of food, a cup, and utensils.a fork and a). The presence is indoors in a restaurant-like which the the elements objects are present and match depicted. any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..42dc6f6b6f46d34749bb3607879cff1dfb427d6a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e2491ed3f536372e6283752425340f61665c3e7df2253c41316c6ba4fb0db657
+size 135583
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..b085b6eafff50b5d29a438a89bf16d1f2db33b9d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a silver SUV on display at a car show. The car is parked on a gray carpet with blue lines, indicating a designated area for the vehicle. The SUV is positioned in the center of the frame, with a red car visible in the background. The car show appears to be indoors, with a large building visible in the background. The style of the video is a straightforward, unedited shot of the car on display, likely taken by a visitor or staff member at the car show. The focus is on the silver SUV, showcasing its design and features."
+    ],
+    "video_ids": [
+        "td9EHnU1uIk_4_0to126"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Silver SUV, Red car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a silver SUV, which is the main subject, and a red car is also visible in the background. Both objects are clearly present and match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a silver SUV, which is the main object, and there red car is also visible in the background. The elements are present present and match the description provided.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..47667235a412913594e80d578db69d6678c8a489
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b6c8a5b8f4045ad5f0682a984b08a395ea9ac08d39502f1c286c3fc972ced6c
+size 77608
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..648b6906c8e254223c4914ad77d99367244a6b8f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated depiction of a garden scene. It features a wooden fence, a metal bucket, and several carrot plants growing in the soil. The garden is set against a backdrop of lush green trees and a fence. The carrot plants are in various stages of growth, with some showing green leaves and others displaying orange roots. The metal bucket is placed near the plants, possibly for watering or collecting produce. The overall style of the video is realistic with a touch of cartoonish elements, giving it a charming and inviting look."
+    ],
+    "video_ids": [
+        "Mu8G3TR_-jw_29_0to120"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Wooden fence, metal bucket, carrot plants? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a wooden fence surrounding the garden, a metal bucket placed in the soil, and multiple carrot plants growing in rows. These core objects are consistently present throughout the video, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a wooden fence, a area area two metal bucket placed on the foreground, and carrot carrot plants growing in the. These elements elements match present present throughout the frames, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5509e627de626da5cfa9832951c5ef9c219f3a85
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba21bf2995ee5dc1360a4461cbde9b3bc67b33c5a021afb0a9913157fa1f4789
+size 141332
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..53dd46ac58f03a1058f56d8b711de1757458f159
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene underwater scene inside an aquarium. The background is a deep, dark blue, providing a stark contrast to the vibrant greenery and colorful fish swimming gracefully through the water. In the foreground, a variety of aquatic plants dominate the view. These plants include broad-leafed species with rich green hues and others with reddish-purple leaves, adding depth and texture to the scene. Among these plants, a tall, spiky plant stands out, its elongated leaves reaching upwards.  As the video progresses, several small fish can be seen swimming around the plants. Their movements are fluid and graceful, creating a dynamic yet peaceful atmosphere. The fish vary in color, with some displaying bright blue and yellow patterns, while others have more subdued tones. They dart between the plants, occasionally pausing to explore their surroundings or interact with each other.  Throughout the video, the camera remains stationary, allowing viewers to fully absorb the tranquil beauty of the underwater environment. The consistent lighting highlights the details of"
+    ],
+    "video_ids": [
+        "5be7af1c41a39525bdd7a22e5b3d303969c179748139fa051a7a1c5e9fd0b38a"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Aquatic plants, colorful fish? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a variety of aquatic plants with different colors and shapes, including green, red, and brown foliage, which fills the bottom and mid-levels of the tank. Several colorful fish, including neon tetras and other small fish, are visible swimming in the water. These elements align with the 'Object(s)' condition of aquatic plants and colorful fish.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts aquatic vibrant of aquatic plants with vibrant colors and shapes, which green, red, and yellow hues. which ful the majority part middle-section of of the frame. Additionally colorful fish, including blue blueras and possibly small,, are swimming swimming around the background, The elements align with the descriptionObject(s)' condition of the plants and colorful fish,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..46325275675ed5635905549ac3f28c569aa77737
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:adf4a79f49a35b05290be2e57f7ad1d74105576faf1469ffbb5067ba4f163737
+size 197283
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..b8dd9c67eed91a2a29dc902bb919da824a5ecde3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a blue shirt standing in a modern living room with large windows. He is holding a smartphone in his hands, looking at the screen with a smile on his face. The room is well-lit with natural light coming through the windows. There is a desk with a computer monitor in the background, and a comfortable couch is visible to the left. The man appears to be enjoying his time, possibly browsing social media or texting someone. The overall style of the video is casual and relaxed, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "R3WW7mnWzkk_24_17to146"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue shirt, a smartphone, a desk, a computer monitor, a couch? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a blue shirt, holding a smartphone, standing in front of a desk with a computer monitor, and near a couch. All specified objects are present and correctly identified in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing a blue shirt, holding a smartphone, and in a of a desk with a computer monitor, and there a couch. The the elements are present and match described, the scene.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..91d5d5249a0fa346f0606e5670e06dd87608589e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1298-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d4c2974e9bcc25cd1ef1fe62a8b726cef157f2384d3d696c0b581c13113e2c35
+size 167142
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..e4e8a20ad731f0f78c174fc76d8f85e36fb38ef2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a 3D animated character, a young girl with vibrant red hair and a strawberry-themed outfit. She is standing in a lush, green garden filled with strawberry plants. The girl is wearing a pink hat with a green brim, a white t-shirt with a strawberry design, and a pink skirt with strawberry patterns. She has a surprised expression on her face, as if she has just discovered something interesting. The garden is filled with ripe strawberries, and the girl seems to be enjoying her time in the garden. The overall style of the video is bright and cheerful, with a focus on the girl and her surroundings."
+    ],
+    "video_ids": [
+        "C-SMUAKmt9A_12_0to172"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young girl with vibrant red hair and a strawberry-themed outfit.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young girl with vibrant red hair and a strawberry-themed outfit, including a strawberry-shaped hat and a pink dress with strawberry patterns. These elements match the core description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young girl with vibrant red hair, a strawberry-themed outfit. which a pink design design and a shirt skirt with strawberry patterns. The elements clearly the description description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bbe9821c7fab7856df1f89742b260a3d5f50bafe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:edb4d83a18e34a55e627b3796a0621eec960562939dc7689634370d2bff7bdcb
+size 231041
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..df7914a1e120fb83f69117853fe154154d85ce3a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a car engine, focusing on the various components and parts. The style of the video is informative, likely aimed at providing a detailed look at the engine's components for educational or diagnostic purposes. The engine is complex, with numerous hoses, wires, and other parts visible. The colors are muted, with the engine's metallic components and the hoses' colors being the most prominent. The video is likely shot from a high angle, providing a comprehensive view of the engine's layout. The focus is on the engine's functionality and design, rather than aesthetics."
+    ],
+    "video_ids": [
+        "AwIBeySukbE_21_21to186"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Car engine, hoses, wires, and other parts.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a car engine with visible hoses, wires, and various engine components. The engine bay is detailed and matches the description, with no conflicting elements present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a car engine with various hoses and wires, and other mechanical parts. The presence is is open, includes the description of indicating no conflicting elements present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fd1a91414ca24543e0a11276c55aaede99bce8f2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6a3d292f7fb89019ded7f35529ed60ab729691d1156219d0e2410399fbe3f2b
+size 206466
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..e91a8e814c3ac176d83bb52d284abd7a2bba7a4f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a body of water, likely a lake or pond, reflecting the surrounding trees and sky. The water is calm, with gentle ripples moving across its surface, suggesting a light breeze. The reflection of the trees is prominent, showing a mix of green and orange foliage, indicating a transition between seasons, possibly autumn. The sky above is overcast, with a soft, diffused light that enhances the tranquil atmosphere. A small bird can be seen floating on the water's surface, adding a dynamic element to the otherwise still scene. The camera remains stationary throughout the sequence, focusing on the interplay between the water, the reflections, and the subtle movement of the bird. The overall mood conveyed by the video is peaceful and contemplative, inviting viewers to appreciate the quiet beauty of nature."
+    ],
+    "video_ids": [
+        "833bca15c60b0cc61847a47d8e2f148effe0a40c08df26a95304d36c402f3576"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small bird? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small white object floating on the water, which is consistent with the description of a small bird. Although the object is not clearly identifiable as a bird due to the reflection and distance, its size and movement are plausible for a bird, and there is no clear contradiction with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a small bird bird floating on the water, which could consistent with the description of a small bird. The the object is not clearly identifiable as a bird, to the lack and distance, the size and position suggest consistent for a small. especially it are no conflicting contradiction to the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..120f067f6fcfdc1cc7d374745c45317a3f813a9b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:edf951bf0369c4694f8cbe3f161ec0e52ba48596a3804d99b5b51949cfb8330a
+size 230233
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..b44b207f8becb6dae4a000efd15b1e50a2a62965
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a kitchen, waving his hand in a friendly manner. He is wearing a black t-shirt with the words \"Jesus Christ\" on it, and a white baseball cap. The kitchen is white, with white cabinets and a white refrigerator. The man is standing in front of a window, which is covered with blinds. The window lets in natural light, illuminating the room. The man appears to be in a good mood, as he is smiling and waving. The overall style of the video is casual and friendly."
+    ],
+    "video_ids": [
+        "RMDdhkkaDWU_4_26to162"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black t-shirt with 'Jesus Christ' on it, and a white baseball cap.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black t-shirt with 'Jesus Christ' printed on it and a white baseball cap, which matches the core description. The background and other elements do not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black t-shirt with theJe Christ' written on it and a white baseball cap. which matches the description description provided The presence and additional elements in not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ad70a1ff0f5ff2998d6fb61ff343019a8bae6c06
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d095f4778e354cf84d8df7bdfc25dc22ac6e29df87b0db706a73eb735ce96115
+size 164720
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..96b4ad8a209d143cf32685bbfba62b4ad9ef7ab8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a playful scene featuring a small yellow bird interacting with a tennis ball on a light wooden floor. The bird, with its bright yellow feathers and small size, is seen perched atop the tennis ball, which is positioned near the center of the frame. As the video progresses, the bird begins to move around the ball, occasionally adjusting its position and orientation. The camera remains stationary throughout, focusing on the bird's movements and interactions with the ball. The lighting in the room is bright, casting soft shadows on the floor, and the background includes a glimpse of furniture legs, suggesting an indoor setting. The bird's actions are gentle and curious, showcasing its playful nature as it explores the ball. The overall atmosphere is calm and serene, highlighting the bird's interaction with its environment."
+    ],
+    "video_ids": [
+        "851463766db89af0ef0587e02c009140245c8ce9a5ca0e486848569aad129235"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small yellow bird, a tennis ball? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a small yellow bird (a budgerigar) interacting with a tennis ball on a wooden floor. The bird is seen perching on the ball, pecking at it, and moving around it, which aligns with the described objects. No elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a small yellow bird perlikely pargerigar) per with a tennis ball. a wooden floor. The bird is per perched on the ball, movingcking at it, and moving around it, which aligns with the description objects. The additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6aa37292b3b7f5d2aa439fe4626ee72b039873f0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67acb51a7cdb027e37b57b88fc82fd4bbab0578f2d85f7ede6f2e7fd546a1c8f
+size 149836
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..e0aaa7e3175f5b763e6db80b5f855a2fc83f260a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a kitchen, preparing a meal. He is wearing a black apron and glasses, and he is smiling as he cooks. The kitchen is well-equipped with various utensils and appliances, including a blender, a toaster, and a coffee maker. The man is using a wok to cook, and he is stirring the ingredients with a spoon. The kitchen has a red wall, and there is a sign that says \"SUBSCRIBE LIKE COMMENT!\" on the wall. The man seems to be enjoying his cooking, and he is likely sharing his cooking experience with an audience. The video is likely a cooking tutorial or a cooking show, where the man is demonstrating how to prepare a meal."
+    ],
+    "video_ids": [
+        "OyfzqI74uAA_4_0to138"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black apron and glasses, a wok, a spoon, and a sign saying 'SUBSCRIBE LIKE COMMENT!'? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black apron and glasses, holding a wok, and there is a visible sign in the background that says 'SUBSCRIBE LIKE COMMENT!'. While a spoon is not clearly visible in the provided frames, the core elements described are present and consistent with the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black apron and glasses, actively a wok with and using is a spoon spoon that the background that says 'SUBSCRIBE LIKE COMMENT!'. The the spoon is not explicitly visible in the video frames, the man elements of in present, match with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..75b3b55c8e2743840dcde0bc689ea4acc7d53428
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d970d7bd2a76fc7909bb5bc32c45597cfbdaedc1dc9635c25d4dd0f112ae0ce0
+size 213013
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..953e113f166b41583ddaabc037a19dc240754c02
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a vibrant and colorful scene from what appears to be an animated or themed attraction, possibly part of a theme park or a decorative display. The setting is a nighttime environment illuminated by various lights and projections.  Initially, the camera focuses on a wall adorned with whimsical, stylized floral designs and trees, featuring bright colors like red, yellow, and green. A large, stylized sun with Chinese characters is prominently displayed, adding a cultural touch to the scene. The camera then pans left, revealing a stage-like area with a blue and white color scheme, reminiscent of an icy landscape. This area is decorated with snowflakes and features animated figures, including a character resembling a rabbit, sitting atop a snowy structure. The background includes stylized ice formations and a castle-like structure, enhancing the wintry theme.  As the camera continues to pan, it captures more of the icy landscape, showing additional animated figures and decorations. The water in the foreground reflects the lights and colors, creating a"
+    ],
+    "video_ids": [
+        "4238a6d7c735f9079ee46f62a983f8fca2cd81d713b8cf20c30c344a0970db19"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ['Wall with whimsical, stylized floral designs and trees', 'Large, stylized sun with Chinese characters', 'Stage-like area with blue and white color scheme, decorated with snowflakes', 'Animated rabbit character on a snowy structure', 'Stylized ice formations', 'Castle-like structure', 'Water reflecting lights and colors']? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a wall with whimsical, stylized floral designs and trees, a large stylized sun with Chinese characters, a stage-like area with blue and white color scheme decorated with snowflakes, an animated rabbit character on a snowy structure, stylized ice formations, a castle-like structure, and water reflecting lights and colors. All these elements are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features fulfills the 'Object(s)' condition as showcasing depicting a ' with whimsical, stylized floral designs and trees, a large,ized sun with Chinese characters, a stage-like area with a and white color scheme decorated with snowflakes, an animated rabbit character on a snowy structure, stylized ice formations, and castle-like structure, and water reflecting lights and colors. The these elements are present in contribute with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..05e34aa485c4abfe4b99644436dedfc866995f74
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b094d904b2d99bec25a47565ce451754364638677a9f7247afd5dd22236002c
+size 280511
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..02a04d1bf721015d2ac4037754890ba402fba2b6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a basketball player in a green jersey with the number 11 and the word \"BOSTON\" on it. The player is seen in three different positions, each time looking up towards the sky with a focused expression. The background is a blurred basketball court with a hoop visible in the first frame. The style of the video is a slow-motion montage, capturing the player's intense concentration and the dynamic movement of the game."
+    ],
+    "video_ids": [
+        "WgRcNDMRxX8_11_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A basketball player in a green jersey with the number 11 and the word 'BOSTON'.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a basketball player wearing a green jersey with the word 'BOSTON' and the number '11' clearly visible. The player's attire and the visible text match the description exactly, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a basketball player wearing a green jersey with the number 'BOSTON' and the number 11' prominently visible. The player is jersey and the setting text on the description provided. fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5c9d146e8cac82cde7546bf5e619ee06792780ad
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1299-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b20a79219e4d897765360bc02b7f23791e69d769b9944f7865ad12b960a3010
+size 253167
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..3bb98a31bb76bbd5722873b20c6f11c5be39e6ba
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two young children are seated at a dining table, enjoying a meal of cereal. The boy, dressed in a yellow shirt, is on the left side of the frame, while the girl, wearing a pink shirt, is on the right. They are both using spoons to eat their cereal, which is served in white bowls. The table is set with a white tablecloth, and there are two bottles of milk on the table, one near each child. The children are seated on gray chairs, and the scene takes place in a room with a white wall in the background. The video captures a simple, everyday moment of childhood, with the children enjoying their meal together."
+    ],
+    "video_ids": [
+        "35CYrD03DJo_38_0to189"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two children, a boy in a yellow shirt and a girl in a pink shirt, two spoons, two white bowls of cereal, two bottles of milk, and two gray chairs.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two children, a boy in a yellow shirt and a girl in a pink shirt (the girl's shirt has a pink 'Barbie' design), sitting at a table with two white bowls of cereal, two spoons, and two bottles of milk. They are seated on two gray chairs. All specified objects are present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two children, a boy in a yellow shirt and a girl in a pink shirt,matching ' is shirt is a pinkishPbie' logo), sitting at a table with two white bowls of cereal, two bottlesons, and two bottles of milk. They are eating on gray gray chairs. The the objects are present and correctly the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ad6da95fb1f97ad9216aa87d78963d40ef00c77b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63ef3e1272c3bf39f7e06680ef4d677d8948582e50f3b14e27166af02f0b6591
+size 93702
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..b4ddd76fe3087e24e8b28687fd27e023a7d1bf6b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a luxurious motor yacht speeding across the open ocean. The yacht, painted in white, is sleek and modern, with a large cabin and a spacious deck. The boat is moving at a high speed, leaving a trail of white foam in its wake. The ocean around the yacht is a deep blue, with small waves breaking around the boat. The sky above is a clear blue, with a few clouds scattered in the distance. The yacht is the only object in the video, and it is moving towards the right side of the frame. The video is shot from a high angle, giving a bird's eye view of the yacht and the ocean. The style of the video is dynamic and action-packed, capturing the thrill of speed and the beauty of the ocean."
+    ],
+    "video_ids": [
+        "AnOKqFaCtt4_65_0to122"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Luxurious motor yacht? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large, white motor yacht cruising at speed on open water, leaving a prominent wake. Its design, size, and amenities (such as outdoor seating and a flybridge) are consistent with a luxurious motor yacht. The setting and presentation align with the description of a luxurious vessel.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a large, white motor yacht with on high on the water. which a trail wake behind The sleek and including, and the (such as the seating and a canopybridge) are consistent with a luxurious motor yacht. The setting and the align with the description of a high motor.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d47e1fb5ef3b655580d81b1f0c7f8695dbcd80cd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a671f6176633dc92ce11a9d5e7840c3d5b6d2afca81d03e0a60fed94e122ab7
+size 225719
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..acf82c1f795d361911cd0bb76e72ff54d8753a7c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a blue muscle car parked on a concrete surface, possibly a parking lot or a street. The car is a classic model, with a sleek design and a shiny finish. The wheels are large and black, with a distinctive design that adds to the car's sporty look. The car is parked in front of a building, which appears to be a warehouse or a factory. The building has a large, flat roof and is made of concrete or metal. The sky is clear and blue, suggesting that the weather is sunny and pleasant. The car is the main focus of the video, and it is shown from different angles and distances, highlighting its design and color. The video has a vintage feel, with a focus on the car and its surroundings. The car's design and color are emphasized, and the background is kept simple and uncluttered. The video is likely intended to showcase the car's design and appeal to car enthusiasts."
+    ],
+    "video_ids": [
+        "gh6Leo6y1Oc_8_42to187"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue muscle car, large black wheels, a building (warehouse or factory)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a blue muscle car with large black wheels, parked in front of a building that appears to be a warehouse or factory. The car's color, wheel size, and the building's industrial appearance match the description. The video does not contradict the core elements specified.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a blue muscle car with large black wheels, parked in front of a building that resembles to be a warehouse or factory. The car's design and the design, and the building's structure look align the description provided The presence does not contradict the given elements of in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..969a26c61a68501c09eb58a751c596469c00838b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5877165b8db6736b493c95475fd6a34f3603e893b23ef52702ef0a03d06fd64
+size 180496
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..0d9538b6f1a01f8f945ac793593d1f443511508e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a stylized, animated depiction of a group of superheroes. The characters are drawn in a dynamic, comic book style with vibrant colors and detailed costumes. The characters are standing in a line, each with a unique pose and expression, suggesting they are ready for action. The background is a dramatic, shattered wall with debris flying around, indicating a recent battle or conflict. The overall style of the video is energetic and heroic, with a focus on the characters' strength and readiness to face any challenge."
+    ],
+    "video_ids": [
+        "M5xs_BFYouc_4_0to173"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A group of superheroes standing in a line, each with a unique pose and expression.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group of superheroes standing in a line, each with a unique pose and expression. Characters like Superman, Green Lantern, The Flash, and others are depicted with distinct stances and facial expressions, fulfilling the 'Object(s)' condition. Additional elements like lightning effects and a speech bubble do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group of three standing in a line, each with a unique pose and expression. The are the, Flash Lantern, and Flash, and others are depicted in distinct stances and expressions expressions, which the 'Object(s)' condition.\"\n The elements like the and and a city bubble do not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6c99fa2625545aff22066f0d5e317625460b7c56
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ecca43c93d9edd7fd6fffcd7b6bb1e89f4968d50dad3657c020fb9ca5e93f880
+size 316086
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..cb4cb4ca5da98bb2c749ea5338b28eb32840c516
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a man in a black suit and a bald head. He is sitting in front of a large screen displaying a cityscape and a man in a suit. The man on the screen is wearing a blue tie. The news segment is titled \"U.S. Asia Pivot Strategy\" and it discusses the U.S. pushing an agreement on TPP. The video is from CCTV America, as indicated by the logo in the corner. The man in the black suit is the main focus of the video, and he appears to be speaking about the topic at hand. The cityscape in the background adds a sense of location to the segment. The overall style of the video is professional and informative, typical of a news broadcast."
+    ],
+    "video_ids": [
+        "WskceRIK3kc_3_799to997"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a black suit, bald head, large screen, cityscape, man in a suit with a blue tie? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man in a black suit seated in front of a large screen displaying a cityscape. On the screen, there is also an image of a man in a suit with a blue tie. All elements described in the 'Object(s)' condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man man in a black suit sitting in front of a large screen displaying a cityscape. There the screen, there is a a image of a man in a suit with a blue tie. The elements in in the questionObject(s)' condition are present in match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..30b562822b26b4707dd2d76af17eb136e01424fd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:93d5e4e7f31f28116559cb6dceba4ef339acfa8a0caca9aebdfb579183053a11
+size 70759
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..03b5dd1a366060fdf191b6c1e1712650accee092
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a hand using a piping bag to decorate a yellow star-shaped cookie. The hand is carefully applying black icing to the cookie, creating a heart shape in the center. The cookie is placed on a white surface, and the focus is on the intricate details of the icing work. The style of the video is a simple, yet detailed demonstration of cookie decorating techniques. The video captures the precision and skill involved in creating beautiful cookie designs."
+    ],
+    "video_ids": [
+        "yE0PkQ1ICBo_22_139to260"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Hand, piping bag, yellow star-shaped cookie, black icing? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a hand holding a piping bag, applying black icing to a yellow star-shaped cookie. All the specified objects are present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a hand holding a piping bag, which black icing onto a yellow star-shaped cookie. The the elements objects are present and the depicted in the video,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..656cff53b8b30197710946c3fb5bbad577c1292f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc12987caa455d8ee78663c7b80adc3cd48dc2facde7249221e303feb6b23cf1
+size 93850
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..21a1a90088f872f533aa6a66a7647852295e94aa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young woman with blonde hair, wearing a white and purple patterned top, sitting on a wooden deck surrounded by lush green trees. She is holding a small object in her hands, which she appears to be examining closely. The woman's expression is one of concentration and curiosity, suggesting she is deeply engaged with the object. The setting is serene and natural, with the deck providing a comfortable space for the woman to sit and contemplate her surroundings. The overall style of the video is casual and relaxed, capturing a moment of quiet introspection in a beautiful outdoor setting."
+    ],
+    "video_ids": [
+        "QkP1MELXYso_47_24to192"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman with blonde hair, wearing a white and purple patterned top, and holding a small object in her hands.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman with blonde hair wearing a white and purple patterned top, and she is holding a small object (which appears to be a folded item or cloth) in her hands. These elements match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman with blonde hair, a white and purple patterned top. and she is holding a small object inwhich appears to be a pen piece, a) in her hands. The elements match the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6f53af4fc2947b50cf4223da239d05fd45b6c5fe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0a724bdfb9cee7158e9be4db1fdb3a009c08bbd348dfa05e96b2ab668bfe592
+size 189976
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..e8e520ead48249558483522712cb7ee2a811dde7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit and tie, sitting in front of a green and blue background with the Fox Business logo. He is identified as Joshua Kendall, the author of \"First Dads\". The man appears to be speaking, possibly discussing his book or related topics. The style of the video is a standard television interview, with the subject seated in front of a backdrop and a logo indicating the network. The focus is on the man and his conversation, with no additional elements or distractions in the frame."
+    ],
+    "video_ids": [
+        "MOUalf_xIZs_1_29to184"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, identified as Joshua Kendall.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a suit and tie, and a lower-third graphic clearly identifies him as Joshua Kendall, author of 'First Dads'. The visual and textual elements align with the described condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit and tie, which the name third graphic includes identifies him as ' Kendall. which of 'El Doine'. The setting elements textual elements in with the description '.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..72b509047e330313630efbb2b03045474cf64cf7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-13-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83be53f5410495026fce6230261f452de4bb407f704a38f8804fd5ef1ed0a696
+size 217719
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..4f160e2527174733e4c253fe070c94a1e31825f8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a van being transformed into a camper van. The first frame shows the van's interior with a white countertop mounted on the wall. The second frame shows the countertop being lowered down to create a table. The third frame shows the countertop fully lowered, revealing a storage compartment underneath. The style of the video is a time-lapse, capturing the transformation of the van's interior in three stages."
+    ],
+    "video_ids": [
+        "d4MqIq63Vb8_4_0to181"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: White countertop? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a white countertop attached to the wall inside the motorhome, supported by golden brackets. This matches the description of a white countertop, and no conflicting elements are present that would contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a white countertop in to the wall in the vehiclehome. which by a legs. This matches the description of a ' countertop as which the other elements are present.\"\n would contradict this description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f48f468e6e5b19156abca53d6538425d0bdde3f8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dcc3169cd2f2a4028c0c73c19b3e1410827967bdaaffee14af82dd47a2cc897d
+size 108739
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..5376a7fd40a7fd5cccb50f0808af265d82314f78
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red Hyundai SUV driving through a parking lot. The car is in motion, moving from the left side of the frame towards the right. The parking lot is filled with various cars, including other red SUVs. The sky is clear and blue, suggesting a sunny day. The style of the video is a straightforward, real-life depiction with no special effects or artistic alterations. The focus is on the red Hyundai SUV, and the video captures its movement and the surrounding environment."
+    ],
+    "video_ids": [
+        "x4d9ZSBDOrA_50_0to176"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red Hyundai SUV? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red Hyundai SUV as the central subject, with clear visibility of its design, color, and branding. Additional cars and background elements are present but do not contradict the core description of the red Hyundai SUV.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a red vehicle SUV, the main object. which its visibility of the design and color, and model. The elements and a elements are present but do not detr the main description of the red Hyundai SUV.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2a87472525b51eed2bdfb09268d7145688322f51
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2cf4ee346e3a605c7a93d8afbcd2be20ce7474824aa6154a421d5cd4c1ebb337
+size 295705
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..43163fdcb3d4933b9f125b4f071bf4b213a7298f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a car's dashboard and center console, focusing on the infotainment system and climate control. The style of the video is a straightforward, unembellished documentation of the car's interior features. The infotainment system displays a menu with options such as \"Auto Tune Response\" and \"Auto Tune Response\", along with a car icon and a red button. The climate control panel has three circular dials, each with a different function, and a digital display showing the temperature setting. The video is likely intended for educational or promotional purposes, showcasing the car's technology and user interface."
+    ],
+    "video_ids": [
+        "m0gdLJGRtsE_16_0to129"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Infotainment system, climate control panel, car icon, red button, three circular dials, digital display showing temperature setting.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows an infotainment system with a digital display, a climate control panel with three circular dials (two large ones and one smaller central dial), and a car icon displayed on the screen. A red button is visible near the bottom right of the screen. The digital display also shows temperature settings and other vehicle information, matching the described elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows an infotainment system with a screen display, a climate control panel with three circular dials,likely for and and one smaller one one), a a car icon. on the screen. There red button is also on the center of of the climate. The digital display also shows a settings, other vehicle information, fulfilling the description elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bde78a225bc972c28edec2db54b68428cf6c7196
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b1478ad7c2fabcbf89874171851900d94ba04aca629b38f4bd652ed32c6cf46
+size 229714
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..b47a78f627af061be64de47d1f9e611fb2908eb7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a bowl of cooked Brussels sprouts on a wooden table. The Brussels sprouts are green and have a slightly charred appearance, indicating they have been roasted. The bowl is glass and has a wide mouth, allowing a clear view of the Brussels sprouts inside. There are also a few cloves of garlic scattered around the bowl, suggesting they were used in the cooking process. The style of the video is simple and straightforward, focusing on the food without any additional context or embellishments. The lighting is bright and even, highlighting the texture and color of the Brussels sprouts. The video does not contain any text or narration."
+    ],
+    "video_ids": [
+        "UGVNqGz0iJg_33_0to139"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bowl of roasted Brussels sprouts, a few cloves of garlic? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a bowl filled with roasted Brussels sprouts, which are visibly caramelized and green. In the background, there are a few cloves of garlic on a wooden surface, matching the described elements. There are no conflicting or misleading elements that contradict the specified objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows shows a bowl filled with green Brussels sprouts, which align green greenized and g, There the background, there are a few cloves of garlic, the wooden surface, which the description elements. The are no additional elements additional elements in contradict the description objects.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c7aedb812a943942d407f9026f1ac6228b9337ac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68da3fa98c337a49190be0b2a226ef72ec54497d636351ca2e8ce092a21f94bf
+size 78017
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..a2ad9b958dd70c1f243bf23f6703cd8d00e1e3e7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a kitchen, preparing food. He is wearing a black apron with the Guinness logo on it. The kitchen is well-equipped with a sink, an oven, and a window. The man is using a knife to cut a piece of food on a cutting board. The video captures the man's actions in a realistic and detailed manner, providing a clear view of the kitchen and the man's activities. The style of the video is realistic and informative, likely aimed at teaching viewers how to prepare a specific dish."
+    ],
+    "video_ids": [
+        "J-o-JTJTa4k_8_0to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a cutting board, a knife, a piece of food, a sink, an oven, and a window.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing an apron, standing at a kitchen counter with a cutting board, a knife, and a piece of food (an onion) on the board. In the background, a sink and an oven are visible, along with a window. All specified objects are present and accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man in an apron, standing at a wooden counter with a cutting board and knife knife, and a piece of food (which un). that the board. There the background, there sink and an oven are visible, and with a window. The the objects are present and the depicted in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dce135ea659c146283df0eb028506b96e29f6ec5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dc01ecb858606c2e0309242a04cd6d116fb9f68798c05eed9525eaac055c69ad
+size 136844
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..da5f408f84d1e1fed6c63e9af112f437fc24d18d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a group of people walking on a circular glass bridge that extends over a deep canyon. The bridge is suspended high above the canyon floor, providing a thrilling view of the vast, rocky terrain below. The group, dressed in casual attire, moves along the bridge, some looking down at the canyon while others gaze out at the surrounding landscape. The bridge's design allows for a clear, unobstructed view of the canyon, creating a sense of awe and wonder. The video is shot from a high angle, emphasizing the height and daring of the bridge and the people traversing it. The overall style of the video is adventurous and exciting, capturing the thrill of exploring nature's grandeur from an unusual vantage point."
+    ],
+    "video_ids": [
+        "IkJNeLo7XTI_25_0to134"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A group of people? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a group of people walking and standing on a glass bridge overlooking a canyon. Their presence and activity match the description of 'A group of people' without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a group of people standing on standing on a glass bridge. a canyon. The presence and the fulfill the description of aA group of people'. in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..58267b6328f4a74af006e411c980f7aa16e99585
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a70f21d9a2d61218b7cdefaf9445fa3ece06807e73b23ed15a065d8d94342887
+size 249726
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..51f5cf425ecd1e59ea83fdb0afa682ca1e862d6f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment from a soccer match between two teams, one wearing white jerseys and the other in yellow. The players are seen walking off the field, possibly after a goal has been scored. The white jerseys are adorned with the logo of a sponsor, while the yellow jerseys are plain. The players are dressed in athletic gear, including shorts and cleats. The field is a vibrant green, contrasting with the players' jerseys. The atmosphere is one of camaraderie and sportsmanship, as the players walk off the field together."
+    ],
+    "video_ids": [
+        "KRP237iFl-4_6_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Players in white and yellow jerseys, field? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows soccer players wearing white jerseys (Real Madrid) and one player in a yellow jersey (opponent goalkeeper), standing on a soccer field. The background includes stadium seating and advertising boards, consistent with a soccer match setting. The core description of players in white and yellow jerseys on a field is accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two players wearing white and andone Madrid) and yellow in in a yellow jersey (likelyponent).). standing on a soccer field. The presence includes a lights and flood boards, which with a soccer field setting. The presence elements of players in white and yellow jerseys on a field is fulfilled represented.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d3206ac0d30ce516abe2754f49f50039020a3f34
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bedd916a7fe1a97c91220207f869f0b78e6632d191754bb46e67eddeb887818a
+size 234953
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..3d685e297399ac8dcdd49a84e3a0aa28d7d2b02c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman stands at a podium, speaking into a microphone. She is dressed in a white polka dot blouse and her hair is styled in loose waves. The podium is made of wood and is positioned in front of a red curtain. The woman appears to be in the middle of her speech, as she gestures with her hands while speaking. The overall style of the video is formal and professional, with the focus being on the woman and her speech."
+    ],
+    "video_ids": [
+        "baw_VHLm6jU_14_17to197"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a podium, and a red curtain.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman standing at a wooden podium, speaking into a microphone, with a red curtain serving as the background. All three specified objects are present and accurately depicted, fulfilling the 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a woman standing behind a podium podium with with into a microphone. with a red curtain in as the backdrop. The the elements elements ( present and match depicted in fulfilling the 'Object(s)' condition.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4f8032327ea43a1f0bf2e12dbb357b96bf761892
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-130-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e82ecfe0475048baa44b8086e0f910662974ff40522ecd93ad847f4e75ec5089
+size 145735
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..5e1a9142ae9e98313906582699543ddff33ef0d6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard, wearing a red plaid shirt and jeans, sitting on a brown leather couch. He is engaged in a conversation, with his mouth open as if he is speaking. The setting appears to be a television studio, as indicated by the presence of a coffee cup and a microphone on a table in front of him. The man's facial expression suggests that he is surprised or shocked by something. The overall style of the video is a standard television interview or talk show segment."
+    ],
+    "video_ids": [
+        "12AznytyQWU_7_16to179"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard, wearing a red plaid shirt and jeans.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard, wearing a red plaid shirt and jeans, sitting on a couch. The description matches the core visual elements of the subject in the video without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard, wearing a red plaid shirt and jeans, sitting in a brown. The description matches the visual elements elements of the video in the video.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e9fd692b2cb26bced3d2aa2bb9b43a623d42af14
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed65fc5943921b16ef127d055ef27fac354b540daf12def5ac5c551c4c700162
+size 343925
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..b6720909c4038678a92c4110eef5d700f0783b00
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a jubilant moment at a football game. A football player, wearing a green and yellow uniform, is the center of attention. He is surrounded by a group of people, all dressed in similar attire, indicating they are fans or possibly teammates. The player is holding a football, suggesting he has just scored a touchdown or made a significant play. The crowd is cheering and celebrating, with some people even lifting the player off the ground in a show of excitement and support. The atmosphere is one of triumph and camaraderie, as the player and his supporters share in the joy of the moment."
+    ],
+    "video_ids": [
+        "X6Ec-hZaToA_64_97to233"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A football player in a green and yellow uniform, a crowd of people, a football.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a football player in a green and yellow uniform (Green Bay Packers), a crowd of enthusiastic fans, and a football being held by the player. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a football player in a green and yellow uniform,Green Bay Packers style a crowd of people players, and a football. held by the player. The elements elements of in present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5e0759295fc4a50bf8ab9e147837b812d0104a26
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f9651ee972fac1af292cc6a98f1b036fbf7ee760a1d912246f8532f18481036
+size 358463
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..281feea8fee51d964ac10fd57760ba09f787d54a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of four cupcakes with white frosting and cinnamon sprinkles on top. The cupcakes are placed on a white surface, and there is a glass jar with a metal lid and a spoon next to them. The style of the video is a simple, straightforward food presentation, with a focus on the cupcakes and their details. The lighting is bright and even, highlighting the texture of the frosting and the cinnamon sprinkles. The background is minimalistic, with a plain white surface that contrasts with the colorful cupcakes. The video does not contain any text or additional elements, and the focus is solely on the cupcakes and their presentation."
+    ],
+    "video_ids": [
+        "fr0f9HWWcy0_0_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Four cupcakes with white frosting and cinnamon sprinkles on top, a glass jar with a metal lid, and a spoon.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows four cupcakes with white frosting and cinnamon sprinkles on top, a glass jar with a metal lid, and a spoon. All elements described are clearly visible and match the given condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows four cupcakes with white frosting and cinnamon sprinkles on top, a glass jar with a metal lid, and a spoon. The these match in present visible and match the core description.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8518359cbf38e0481ba9a97b0c5efcc53a960c23
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a5f8c49eefc973e1950bf4a82e83866000b9737f285860b06e2ab3f0be214ea
+size 47079
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..b920978b7a69d8fbf339e5ac40af52774bf8d55a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman in a white dress is seen interacting with a kangaroo in a grassy field. The woman is bending down to pet the kangaroo, which is standing on its hind legs. The kangaroo appears to be comfortable with the woman's touch, indicating a friendly interaction. The field is lush and green, providing a natural and serene backdrop for the scene. The woman's white dress contrasts with the greenery, drawing attention to her and the kangaroo. The overall style of the video is calm and peaceful, capturing a moment of connection between humans and wildlife."
+    ],
+    "video_ids": [
+        "EI7ATIcXZLA_282_0to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a white dress and a kangaroo? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a white outfit (a white sweatshirt and white shorts, which can be considered a white dress-like ensemble) interacting with a kangaroo in a grassy field. The core objects described \u2014 a woman in white and a kangaroo \u2014 are clearly present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a woman in a white dress interactinglikely dress dressshirt and shorts shorts) which could be considered a white dress in attire) interacting with a kangaroo in a lushy area. The woman elements, in a woman in a and a kangaroo \u2014 are present present and match depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0cd51f7219d87f7ecd092249ed32249c4afb0914
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec4685d46e5104eae4ec8cc4b92f6bb18e27ac102ab2382bbadcd87a227804ea
+size 277733
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..a12950359e7114fe0548c03008d0c8f28faa27dc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a thrilling journey through the desert in a green jeep. The driver, dressed in a beige shirt, is seen holding the steering wheel firmly, navigating the vehicle through the sandy terrain. The jeep is equipped with a spare tire mounted on the back, ready for any unexpected situations. The dashboard of the jeep is visible, indicating the vehicle's speed and direction. The vast expanse of the desert stretches out in the background, with sand dunes visible in the distance. The clear blue sky above adds to the serene yet adventurous atmosphere of the scene. The video is a testament to the thrill of off-roading and the beauty of the desert landscape."
+    ],
+    "video_ids": [
+        "DYXrh7mcKtY_67_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Green jeep, spare tire, driver (dressed in a beige shirt), dashboard? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a green jeep with a spare tire mounted on the back, a driver wearing a beige shirt, and a visible dashboard with gauges. All elements described in the condition are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person jeep with a spare tire mounted on the back, a driver wearing a beige shirt, and a dashboard dashboard. auges and The elements in in the condition are present in match with the video content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d5327e6513e56f828b76353a5d0043808b6be150
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d379cb4d5be524dbff51f67f998fbb8c90c85c852b34ef36807b8c09ef73355b
+size 195096
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..adb9fed6e5d80b1467296ee2da70b419af9b82f1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen interacting with a porcupine in a brick enclosure. The man, dressed in a white tank top and a baseball cap, is seen reaching out to the porcupine, which is situated on a bed of straw. The porcupine, with its spiky quills, appears to be comfortable with the man's presence. The brick enclosure provides a simple and uncluttered backdrop for this interaction, allowing the viewer to focus on the man and the porcupine. The overall style of the video is candid and unscripted, capturing a moment of human-animal interaction in a natural setting."
+    ],
+    "video_ids": [
+        "S6alQ23AkWc_69_0to131"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a porcupine, straw, a brick enclosure? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting on the ground next to a porcupine, which is on a bed of straw. They are both inside a brick enclosure. All the specified objects are present and correctly depicted without any contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man interacting in the ground, to a porcupine. which is resting a bed of straw. The are within within a brick enclosure. The the elements elements ( present in the described in any contradictions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0545a4120d117e777198a0dd0f26771159a60cfa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d095f2144310dde1efc8ee5b9b20584a8072b467fdc91b1d57f6e05554432c1d
+size 288683
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..7de004aaf75b37aa33f5a06839d1b7ad8fb1e013
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the majestic golden dome of a mosque, set against the backdrop of a bustling city. The mosque, with its intricate design and golden dome, stands out prominently in the frame. The cityscape in the background is a mix of towering buildings and lush green trees, creating a harmonious blend of urban and natural elements. The video is taken from a high vantage point, providing a bird's eye view of the mosque and the city. The perspective allows for a comprehensive view of the mosque and its surroundings, highlighting the mosque's prominence in the cityscape. The video is a testament to the mosque's architectural beauty and its significance in the city's landscape."
+    ],
+    "video_ids": [
+        "ot7xCM-Oeig_442_0to133"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Mosque with a golden dome, tall buildings, green trees? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a mosque with a golden dome, surrounded by green trees and with tall buildings visible in the distant skyline. These elements align with the description provided, confirming that the video successfully fulfills the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a mosque with a golden dome, which by tall trees and tall tall buildings in in the background background. The elements align with the description provided, making that the video successfully fulfills the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..390562e2bc8a0d17a614735e212cd6eaeab19014
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a38349301c531b224279bf96c43a75331dbdf6373971c8cbc2e387d83208db3c
+size 155559
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..81821c979cd82bbbe97fa4ba59d403b8f3593a4e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a promotional advertisement for a red Porsche sports car. The car is shown in three different angles, highlighting its sleek design and vibrant color. The first angle is a close-up of the car's front, showcasing the iconic Porsche logo and the car's headlights. The second angle is a side view, emphasizing the car's aerodynamic shape and the shiny red paint. The third angle is a rear view, focusing on the car's taillights and the Porsche logo on the trunk. The car is set against a backdrop of a winding road, suggesting the car's performance capabilities. The overall style of the video is sleek and modern, with a focus on the car's design and performance."
+    ],
+    "video_ids": [
+        "voONmgwMpRI_2_0to153"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red Porsche sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red Porsche sports car, identifiable by its distinctive design, headlights, and the Porsche emblem on the hood. The car is the central focus of the video, and no elements contradict the description of it being a red Porsche sports car.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a red Porsche sports car, which by its distinctive design features including, and overall Porsche emblem. the front. The car is depicted central focus of the video, and there additional contradict the description of a being a red Porsche sports car.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c33417b8637b3bead1f3acc47f38070bcc6ef6f4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1300-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5dc68eea78d0d27314c9236c2b67fd83c03f7e5fb2a5ddae35d2e9cc408bd7d
+size 237014
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..9d501192d6461d1a8b73f3176726764ff03c2276
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated sequence set in a basketball arena. The main focus is a basketball player in a white jersey with red accents, standing on the court. The player is looking to the side, possibly at a teammate or opponent. In the background, there is a referee dressed in a black and white striped shirt, standing on the court. The referee is looking towards the player, possibly observing the game. The court is surrounded by a crowd of spectators, who are watching the game intently. The crowd is a mix of people, with some standing and others sitting on benches. The benches are arranged in rows, and the spectators are dressed in a variety of colors. The overall style of the video is realistic, with attention to detail in the characters and the environment. The animation is smooth and fluid, capturing the dynamic nature of the game."
+    ],
+    "video_ids": [
+        "hKCab1hRb5o_4_0to193"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A basketball player in a white jersey with red accents, a referee in a black and white striped shirt, and a crowd of spectators.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a basketball player in a white jersey with red accents (identifiable as a Cleveland Cavaliers player), a referee in a black and white striped shirt, and a crowd of spectators in the background. All core elements described are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a basketball player in a white jersey with red accents,numberifiable by number player Cavaliers jersey based a referee in a black and white striped shirt, and a crowd of spectators in the background. The elements elements of in present, match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7337428cf05d5f39921e518508dfa0cf57b6d938
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c61878df0eb127bc71b513a6755c76c0715cd19fbed20ea711d2e43621b98fee
+size 198579
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..03d3f4d335ea5fad8823395097e51d3564841398
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene journey on a boat across a calm lake. The boat, painted in shades of blue and white, is equipped with a sturdy metal railing for safety. A blue backpack rests on the boat, perhaps filled with essentials for the journey. The lake itself is nestled amidst a lush forest, with tall trees standing tall in the background. The sky above is a clear blue, dotted with fluffy white clouds. The boat's journey across the lake is peaceful and unhurried, allowing the viewer to soak in the tranquility of nature."
+    ],
+    "video_ids": [
+        "Go2uMKRTGXc_6_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Boat, blue backpack? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a boat moving on water with a blue backpack secured on its deck. These objects are central to the scene and are accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a boat with on a, a blue backpack visible on the side. The elements are present to the scene and align consistent represented, any.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..965934173066c13c698d9b5c9696aea24b82ef7f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7637bded8bccc915482b16495df3bb029f3a0a53a7c3c51a466548e5e2fbd9e5
+size 245815
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..7fdf92822498099be8b9d57ad7752b6827e3b982
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man walking past a row of luxury sports cars on display at a car show. The man, dressed in a casual gray t-shirt, is seen from behind as he strolls along the line of cars. The cars, varying in color and design, are parked neatly in a row, showcasing their sleek and aerodynamic shapes. The setting appears to be an outdoor car show, with the cars arranged on a paved surface. The man's casual attire contrasts with the high-end nature of the cars, adding a touch of everyday life to the scene. The video is a snapshot of a moment at a car show, capturing the interest and admiration of the man for the luxury sports cars on display."
+    ],
+    "video_ids": [
+        "EusQLfTIwhk_33_16to178"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man, luxury sports cars? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features multiple luxury sports cars, including recognizable models like Ferraris and Bentleys, parked in a row. Several men are visible walking among or standing near the cars, engaging with each other or observing the vehicles. The scene aligns well with the 'Man, luxury sports cars' condition, with no significant contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a luxury sports cars, which a models like thearis, Lambleys, which in a row. The cars are present, in the near near these cars, which with the other or the the vehicles. The setting iss well with the descriptionObject, luxury sports cars' condition, as the conflicting contradictions or}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..75924bd4328afa81343e9a8315418518336c5f4d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b1e55cff53416d32c6098a653a6aea28129d52de4d1bb5c8d7d98d05a4fdf53
+size 180760
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..36417659709a821a0875b4aaf192c5abb15e73a9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a man in a black suit and a bald head. He is sitting in front of a large screen displaying a cityscape and a man in a suit. The man on the screen is wearing a blue tie. The news segment is titled \"U.S. Asia Pivot Strategy\" and it discusses the U.S. pushing an agreement on TPP. The video is from CCTV America, as indicated by the logo in the corner. The man in the black suit is the main focus of the video, and he appears to be speaking about the topic at hand. The cityscape in the background adds a sense of location to the segment. The overall style of the video is professional and informative, typical of a news broadcast."
+    ],
+    "video_ids": [
+        "WskceRIK3kc_3_799to997"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a black suit, bald head, large screen, cityscape, man in a suit with a blue tie? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man in a black suit seated in front of a large screen displaying a cityscape. On the screen, there is also an image of a man in a suit with a blue tie. All elements described in the 'Object(s)' condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man man in a black suit sitting at front of a large screen displaying a cityscape. There the screen, there is a a image of a man in a suit with a blue tie. The elements in in the questionObject(s)' condition are present in match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..39b48fc39dd8b23bd1d71daf4bb3d1c9d89fae1f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:803a8b8dc8e3beb49e338d896555e274525d01c393a762be5fead4a726c13bfd
+size 111944
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..259e437fcf050472585825ce546b4d0c7e684afd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a chef in a professional kitchen, preparing a meal. The chef is wearing a white chef's coat and is focused on the task at hand. The kitchen is well-equipped with various cooking utensils and ingredients. The chef is seen using a knife to chop vegetables, and a bowl is visible on the counter. The chef is also seen using a blender, suggesting that a smoothie or sauce is being prepared. The overall style of the video is realistic and informative, providing a glimpse into the daily workings of a professional kitchen."
+    ],
+    "video_ids": [
+        "XDvq231VB30_25_24to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chef, knife, vegetables, bowl, blender? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chef in a kitchen, mixing ingredients in a bowl. A blender is visible in the background, and there are vegetables (including colorful chopped ones on a tray) on the counter. All the specified objects are present, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chef in a professional environment actively and in a bowl. The knife is not in the background, suggesting the are various andlikely broccoli bell items) a cutting) and the counter. The the elements objects ( present and and the 'Object(s)' condition.\"\n}<|im_end|>\n addCriterion",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..69de5b18bc6146132444a34795e4832d108d0fc7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87b6269cde1ae9d5c2234274fa10248c0fe3996fa1487aeb26c655b20f7850fc
+size 118744
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..f5eabba662d726e08fc10d797da2d4d93300ad00
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a lively soccer practice session on a large, well-maintained grass field. The field is marked with white lines, indicating the boundaries and goals. Several soccer nets are set up across the field, and players are scattered around, engaged in various drills and exercises. Some players are seen running, while others are practicing their ball control and passing skills. The players are dressed in athletic attire, and the scene is filled with energy and movement. The field is surrounded by a neatly trimmed lawn, and the overall atmosphere is one of organized chaos, typical of a busy soccer practice session."
+    ],
+    "video_ids": [
+        "cPYduVWocSE_3_0to102"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Players, soccer nets? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows multiple players on a soccer field, actively engaged in training or practice, and several soccer nets are visible positioned across the field. These elements align with the 'Object(s)' condition of players and soccer nets.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts multiple soccer and a soccer field, engaging engaged in a or a. and there soccer nets are visible. around the field. The elements align with the 'Players(s)' condition specified the and soccer nets,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..74b3b41c0f3ff03610d93cde16f7b5925c8e6b08
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:642227a8a09404e162b5f679258931f33dbb46d16f27033db87a36d569ed148d
+size 341835
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..5e11428a98013c0990cc69808fd3eea6200bd67b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a close-up of a brown teddy bear with large, round, yellow eyes and a black nose. The teddy bear appears to be looking directly at the camera, giving the impression of a friendly and curious expression. The background is blurred, but it seems to be a natural setting with greenery, suggesting that the teddy bear might be outdoors. The style of the video is simple and straightforward, focusing on the teddy bear's face and expression. The lighting is soft and even, highlighting the teddy bear's features without creating harsh shadows. The overall impression is one of warmth and comfort, typical of a beloved stuffed toy."
+    ],
+    "video_ids": [
+        "XTL37bnlI1w_21_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Brown teddy bear with large, round, yellow eyes and a black nose.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a brown teddy bear with large, round, yellow-rimmed eyes and a black nose, which matches the description. The bear's features are clearly visible and consistent with the specified attributes.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close teddy bear with large, round, yellow eyesimmed eyes and a black nose. which matches the description provided The bear's fur are clearly visible, consistent with the given characteristics.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b7df9acb447745ba32d17f09753131500acfa4ba
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5716be9762fc304e8cd81a6a5430c0b61bcfa4bb463cc477891dfda230dc3de
+size 107406
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..4ee06bd7cdda27bd9b17acf9fe668923f3e2ccfe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of intense emotion and reflection for a soccer player. The player, dressed in a white jersey with black stripes, stands on the field with his eyes closed and hands clasped together in a gesture of prayer or contemplation. The background is a blur of spectators, indicating that the scene takes place in a stadium during a match. The player's focused expression and the blurred crowd in the background create a sense of solitude and introspection amidst the bustling atmosphere of the game. The video is a still image, capturing a single moment in time, and the style is realistic, with a focus on the player's emotional state and the atmosphere of the stadium."
+    ],
+    "video_ids": [
+        "xlAj1My5Rjc_26_0to158"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A soccer player dressed in a white jersey with black stripes, standing with eyes closed and hands clasped together.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a white jersey with black stripes, standing with eyes closed and hands clasped together. The jersey also features the Real Madrid crest and UEFA Champions League patch, which aligns with the description. The player's posture and attire match the specified conditions without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a white jersey with black stripes, standing with eyes closed, hands clasped together. The player and features a number Madrid logo, the Champions League logo, which ares with the description. The player is posture and the match the given '.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8ea247983a9e5148966cfb052151c20059a6bf84
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1301-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2fc46b0faed1ca1c04a715aa1f0732296cc3bf0ceeaf83edf2d64c5b1136150
+size 56209
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..2d9c50e0e57b5d50ffb8e377fbd09dd4bf6e396c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a blue and red soccer jersey, walking on a soccer field. The man is looking to the side, with a focused expression on his face. The jersey has the word \"Rakuten\" written across the front. The background is blurred, but it appears to be a soccer stadium with spectators in the stands. The style of the video is a close-up shot, focusing on the man's face and upper body. The lighting is bright, suggesting it's daytime. The man's posture and expression suggest he is a professional soccer player, and the setting implies he is at a soccer match."
+    ],
+    "video_ids": [
+        "0w-KVk1Z5zg_27_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a blue and red soccer jersey with 'Rakuten' written on it.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue and red striped soccer jersey with the word 'Rakuten' clearly visible across the chest. The jersey also features the FC Barcelona crest and Nike logo, consistent with the description. The background is blurred, focusing attention on the man and his jersey, which matches the specified object condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue and red soccer soccer jersey with ' ' 'Rakuten' prominently visible on the chest. The jersey also features a Nike Barcelona crest and a logo, which with the description. The background appears a, suggesting attention on the man, his attire, which align the ' condition condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c1ec67632faca413960861188904e9333e6bd011
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:228bd37415176ffc1c6015b781eef05735062cef575300c23393eaaa676c8908
+size 376386
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..ce80c4e63f1a27178f348a55d7655984eff042d9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a soccer player in action, wearing a white jersey with a black and white logo on the left chest. The player is seen in three different positions, each time looking up at a soccer ball that is in the air. The first frame shows the player looking up at the ball, which is in the top left corner of the frame. In the second frame, the player has moved slightly to the right and is still looking up at the ball, which is now in the top right corner of the frame. In the third frame, the player has moved even further to the right and is still looking up at the ball, which is now in the top center of the frame. The player's focus and determination are evident in each frame, as he prepares to make a move on the ball. The background is blurred, but it appears to be a soccer field, suggesting that the player is in the middle of a game."
+    ],
+    "video_ids": [
+        "8RtjVuT0V34_53_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A soccer player in a white jersey with a black and white logo on the left chest, a soccer ball in the air.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a white jersey with a black and white logo on the left chest, which matches the description. A soccer ball is also visible in the air, although it is slightly blurred due to motion. There are no elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a white jersey with a black and white logo on the left chest. and matches the description. Additionally soccer ball is also present in the air, fulfilling it appears not distorted, to motion, The are no additional in contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2509196f76782dae67d9f1820d2b7179ba55a122
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6287c60b4bda93a7b89e3b44eb14b50ed382a36a9edcc7cb62957da94f7d3e62
+size 136638
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..2bbcefd05235551cc98905c15b7e0cf5c2d1828c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two women are cooking in a kitchen. The woman on the left is stirring something in a pot, while the woman on the right is holding her head, possibly in frustration or disbelief. The kitchen is well-equipped with various appliances and utensils, including a refrigerator, an oven, a sink, and a cutting board. There are also several bowls and bottles on the counter. The scene suggests that the women are in the middle of preparing a meal, and the woman on the right might be experiencing some difficulty or confusion in the process. The overall style of the video is realistic and candid, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "NH0P-u-8dzc_34_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two women, one stirring something in a pot and the other holding her head.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two women in a kitchen setting. One woman is actively stirring something in a pot on the stove, while the other woman has her hands on her head, appearing distressed or overwhelmed. This matches the described condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two women in a kitchen setting. One woman is actively stirring something in a pot on the stove, which the other woman is her hand on her head, appearing to or confused. The matches the description scenario of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..71a590b7dd2ce755f1094d048920d6f4dc3aab13
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:49f94affa9b344ed0fca132be9012daaf5ae6a18586ef2d1a20abc7b5a8e782a
+size 174807
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..e73c9e1a5311d77c1519434a564db45ac1d7e796
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a suit and tie, sitting in front of a panel of people. He is smiling and appears to be engaged in a conversation or presentation. The panel consists of a diverse group of individuals, including a woman wearing a red and white scarf, a man with glasses, and a woman with long hair. The setting appears to be a professional or formal event, possibly a conference or panel discussion. The man in the suit is the central figure in the video, and his interaction with the panel suggests that he is the speaker or presenter. The overall style of the video is professional and polished, with a focus on the man and his interaction with the panel."
+    ],
+    "video_ids": [
+        "09RzfX0hbdg_0_0to123"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, a woman wearing a red and white scarf, a man with glasses, and a woman with long hair.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a suit and tie (the host), a woman wearing a red and white scarf (in the audience), a man with glasses (in the audience), and a woman with long hair (in the audience). All these individuals are present and identifiable in the frames provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows shows a man in a suit and tie,which main), a woman with a red and white scarf (part the background), a man with glasses (in the audience), and a woman with long hair (in the audience). The these elements match present in match in the video,,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..43c17fd695ceb58f3e3ef60f689bea9951843c1d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d6e96c96033999f95a0ef1af2f6b967553908346dd75e394997d0adf354853b5
+size 146980
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..8d87206cc044e11fac67d1303e92900f90565fb2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a group of cyclists riding along a coastal road. The cyclists, dressed in blue jerseys, are spread out in a line, each one maintaining a safe distance from the next. The road they're on is a two-lane highway, curving gently along the coastline. On one side of the road, the ocean stretches out, its surface a deep blue. On the other side, the land is covered in lush green vegetation, providing a stark contrast to the blue of the ocean. The sky above is a clear blue, with a few clouds scattered here and there. The overall style of the video is a bird's eye view, giving a comprehensive perspective of the cyclists, the road, and the surrounding landscape."
+    ],
+    "video_ids": [
+        "C5NnNPwBCJE_90_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A group of cyclists dressed in blue jerseys? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a group of cyclists riding along a coastal road, and they are all wearing blue jerseys, which matches the description. The aerial perspective provides a clear view of the cyclists and their attire, confirming the object condition is fulfilled.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a group of cyclists riding on a coastal road. and they are dressed wearing blue jerseys. which matches the description of The presence view and a clear view of the cyclists and their attire, confirming that presence condition.\"\n fulfilled.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..34d1c869dc2da66dcdd823105433ddc4b97801d9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:186862f7909b3467543ee934ca132b906517268f8d3b052e4d4e4e828491cc10
+size 198320
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..b2719d8201a7143fcc0e4dddd732fcc9c691e436
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic scene of three sports cars racing down a winding road. The first car, a vibrant green Porsche GT3 RS, leads the pack, its sleek design and powerful engine evident. Following closely behind is an orange sports car, its bright color contrasting with the green of the first car. The third car, a red sports car, is slightly behind the orange car, adding to the sense of speed and competition. The road they're on is curvy and appears to be in a rural or less populated area, with trees and a power line visible in the background. The sky above is cloudy, suggesting an overcast day. The overall style of the video is dynamic and action-packed, capturing the thrill of high-speed racing."
+    ],
+    "video_ids": [
+        "rU8zy4071HM_13_0to119"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three sports cars: a vibrant green Porsche GT3 RS, an orange sports car, and a red sports car.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows three sports cars in a row: a vibrant green Porsche GT3 RS (identifiable by its model name and color), an orange sports car (likely a Lamborghini Gallardo, based on its design), and a red sports car (likely a Ferrari, based on its shape and color). All three vehicles match the description provided, and no elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows three sports cars: motion sequence: a vibrant green Porsche GT3 RS,frontifiable by its design and and design), an orange sports car,which a Lamborghini basedardo, based on its design and and a red sports car (likely a Ferrari  based on its design and color). The three cars are the description provided, and there additional contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5e859628280f3171a284ce5afa7c7983af5a195a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b21dc59c7ecedf3a9327b4e3101d834a5b698bea8542aa02b38c77467bf6cb49
+size 212786
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..4f312eb5891fac7da8f37d5d2599d036da5a37d6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young woman with long brown hair and blue eyes. She is smiling and looking directly at the camera. She is wearing a sleeveless top and has a black strap over her shoulder, possibly from a bag or a camera. The background consists of a metal fence and trees, suggesting an outdoor setting. The lighting is natural, indicating that the video was likely taken during the day. The style of the video is casual and candid, capturing a moment of the woman's life."
+    ],
+    "video_ids": [
+        "B1s11ZvZC6E_20_97to255"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman with long brown hair and blue eyes, wearing a sleeveless top and a black strap over her shoulder.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman with long brown hair and blue eyes, wearing a sleeveless top. A black strap is visible over her shoulder, consistent with the description. The core elements of the description are accurately represented, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman with long brown hair and blue eyes, wearing a sleeveless top and There black strap, visible over her shoulder, which with the description. The setting elements of the description are accurately represented in and there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b3dc01e308048bd96e3bdea21ff77124e4c133fa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:baba149e06bf4d3fd19430a1399cc30a502b923b4cd07fc8216d753a55888eee
+size 133186
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..cd818aab8be8730a68890135df1c27ed2c72f1ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man with curly red hair, wearing a blue jacket, holding a gold medal with a blue ribbon. He is smiling and appears to be in a celebratory mood. The background is a blue gradient, which suggests that the video might be from a news segment or a television show. The man's expression and the medal he is holding indicate that he has achieved something significant, possibly related to sports or an achievement in his field. The style of the video is straightforward and journalistic, with a focus on the man and his accomplishment."
+    ],
+    "video_ids": [
+        "LsVhSmVHwx8_7_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man with curly red hair, blue jacket, gold medal with blue ribbon, smiling expression.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with curly brown hair, wearing a blue jacket, holding a gold medal with a blue ribbon. He has an open-mouthed expression that can be interpreted as a joyful or celebratory smile. While the hair color is described as red in the prompt, it appears more brown in the video, which is a minor discrepancy. The core elements \u2014 man, curly hair, blue jacket, gold medal with blue ribbon, and smiling expression \u2014 are all present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man with curly red hair, wearing a blue jacket, and a gold medal with a blue ribbon. He is a expression smileed smile that could be interpreted as a smile or smilingatory smile. The the hair color is not as red, the question, it is to brown in the video, which might a reasonable discrepancy but However overall elements of the, curly hair, blue jacket, gold medal, blue ribbon, and smiling expression \u2014 are all present and match the description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b4bef7df51bc901dc000e0d11e210415b8c4de48
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1302-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f3c66a47871a28af657347c6d07d696d7ce38dfee36364a13da233655afa484
+size 226611
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..c0f629689d08b8c4dbcf3c2e5c898e2612047cb6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the dynamic movement of a silver Ford Mustang GT car on a winding road. The car, with its black and white stripes, is seen from the rear, showcasing its iconic design and the vibrant red taillights. The car's license plate reads \"Ford Performance\", indicating its high-performance capabilities. The car's black exhaust pipes are visible, hinting at the powerful engine beneath. The road itself is a blur, emphasizing the car's speed and agility. The background features a desert landscape with mountains in the distance, adding a sense of vastness and freedom to the scene. The overall style of the video is dynamic and energetic, capturing the thrill of driving a high-performance car on an open road."
+    ],
+    "video_ids": [
+        "BL1eyhNVloI_29_122to321"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Silver Ford Mustang GT car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a silver Ford Mustang GT car from the rear, with distinctive taillights, a Ford Performance badge, and a black spoiler, all consistent with the description. The car is in motion on a road, which does not contradict the object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a silver Ford Mustang GT car from a rear view which a featuresillights and a black emblem badge, and a license racing with which of with the description of The car is driving motion on a desert with which is not contradict the core condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8b352ed9010057b46dc74100b90bac47cbef710f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95a489ce17c385fd4568e91780633450ebbe53f370cbbffa874ff08952f8fb2f
+size 153605
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..faddbb12d5dd2b819b269482f6c5058e33b150bb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up shot of a bird's beak as it drinks water from a stream. The bird's beak is positioned at the edge of the frame, with water flowing out of its beak in a thin stream. The background is dark, creating a stark contrast with the bright, sparkling droplets of water that are scattered around the bird's beak. These droplets appear to be illuminated by a light source, giving them a glowing appearance. The bird's feathers are visible, with some parts reflecting light, adding a shimmering effect to the scene. The camera remains stationary throughout the sequence, focusing on the bird's beak and the water droplets. The overall atmosphere is serene and natural, emphasizing the beauty of the bird's feeding behavior and the intricate details of its plumage."
+    ],
+    "video_ids": [
+        "dc51cd64e0929007673e00e8b0d08cb32151f9eb1f0edea387940faa21ac6a9a"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bird's beak, water droplets, bird's feathers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a bird's beak interacting with water droplets, and the bird's feathers are visible, particularly around the head and neck area. The focus and composition emphasize these elements, fulfilling the 'Object(s)' condition as described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a bird's beak, with water,plets, which the bird's feathers are visible in fulfilling the the be and neck area. The presence is detail of these elements, fulfilling the 'Object(s)' condition as described.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b34023b83362668d9ea2c895c397f83a750b5e9c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f20eb3cedd5ab673b3ff5780bfe1228c37e2888447e66d0d2381cc16cc7435d
+size 216856
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..1d770fec4701047b5c704f7969c52a1de3409543
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a kitchen, engaging in a cooking demonstration. She is standing in front of a stove, which has a pot on it. The kitchen is well-equipped, featuring wooden cabinets and a coffee sign on the wall. The woman is holding a spoon in her hand, indicating that she is in the process of cooking. The overall style of the video suggests that it is a cooking tutorial or a cooking show. The woman appears to be the main subject of the video, with the kitchen serving as the backdrop for her demonstration. The video likely includes a combination of visuals and audio, with the woman providing instructions or commentary as she cooks."
+    ],
+    "video_ids": [
+        "Uoykb-j-mcE_2_116to261"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a stove, a pot, a spoon? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman standing in a kitchen, a stove with a pot on it, and a spoon is visible in the utensil holder in the background. All core objects mentioned in the condition are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a woman in in a kitchen, stirring stove with a pot on it, and a spoon. visible in the potil holder next the foreground. The the elements are in the condition are present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b5870b59b6efec68eda9fa5a1881de8d706a39a2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60329335bdbd5a4929d43bedcf49f9d3013de4acf71dcb07ebeab6fbded4caad
+size 127964
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..e3bb8d0293414366a920d09c067b55cd0ce070b6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen sitting in front of a black background with the title \"SOLO: A STAR WARS STORY\" displayed in yellow. She is wearing a black and white polka dot blouse and has her blonde hair styled in an updo. Her hand is raised, and she is gesturing with her fingers. The woman appears to be engaged in a conversation or interview, as suggested by her expressive hand movements and the context of the Star Wars film. The overall style of the video is a mix of casual and professional, with the woman's attire and the Star Wars branding indicating a connection to the film industry."
+    ],
+    "video_ids": [
+        "0xReQ1uExHQ_26_120to245"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a black and white polka dot blouse, blonde hair styled in an updo, a raised hand with finger gestures.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair styled in an updo, wearing a black and white polka dot blouse. She is raising her hand with finger gestures, which matches the description. The presence of the 'SOLO: A STAR WARS STORY' logo in the background does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a woman wearing blonde hair, in a updo, wearing a black and white polka dot blouse. She is gest her hands with finger gestures, which align the description. The background of the logoSUL' AS STAR WARS STORY' logo in the background does not contradict the core description of}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..93c2a5c86a676a074f21de6f12993acd2cde044f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9dc35645c2727a2993030250098b89a0f1a8c0f0c90a88eb05b5adce0097cab0
+size 321281
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..090c2011d5ee5d2ca725ee2428cab26adeb980e5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dramatic scene of firefighters in action. In the first frame, a red fire truck is parked on the side of a road, with a hose attached to its side. The firefighters, dressed in yellow suits, are seen walking towards the truck. In the second frame, the firefighters have reached the truck and are in the process of connecting the hose to it. The third frame shows the firefighters in action, spraying water from the hose onto a large fire that is engulfing a forest. The fire is intense, with flames reaching high into the sky. The firefighters are working tirelessly to extinguish the fire, their yellow suits standing out against the backdrop of the burning forest. The video is a powerful depiction of the bravery and dedication of firefighters in the face of a natural disaster."
+    ],
+    "video_ids": [
+        "wJYQMIcY3YQ_1_0to140"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red fire truck, firefighters in yellow suits, flames? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red fire truck, two firefighters in yellow suits, and intense flames in the background, matching the described 'Object(s)' condition. The scene is consistent with firefighting operations in a forest fire environment.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a red fire truck, firefighters firefighters in yellow suits, and flames flames. the background. which the description elementsObject(s)' condition. The sequence depicts consistent with theing activities, a forested setting.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ad57f0b2b172c58a0cad28a29d4c0334505f5a1b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:883ac44fbd42a7f2f62d5152d12ff7f04528bcf6b394db376324ecb74728b9ad
+size 260391
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..0f3e982cac0ad21b8066889b1d069a915cdad24f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young man with a beard and blue eyes, who is smiling and appears to be in a good mood. He is wearing a gray sweater and is positioned in front of a microphone, suggesting that he might be recording a podcast or a video. The background is black, which puts the focus on the man and the microphone. The style of the video is simple and straightforward, with no additional elements or distractions. The man's expression and the microphone suggest that he is engaging with an audience, possibly sharing his thoughts or opinions on a particular topic. The overall impression is that of a casual and friendly conversation."
+    ],
+    "video_ids": [
+        "FkLJ45Pt-mY_0_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man with a beard and blue eyes, wearing a gray sweater, standing in front of a microphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man with a beard and blue eyes, wearing a gray sweater, standing in front of a microphone. The background is black, which isolates the subject and emphasizes the described elements without contradiction. The microphone is clearly visible and positioned in front of him, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man with a beard and blue eyes, wearing a gray sweater, standing in front of a microphone. The description is dark, and isates the subject and the the microphone elements. any.\"\n The subject is clearly visible, positioned in front of the, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0c62abdeb2bd9599b70e8c330c7f42bd42aa8ba3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d13082c21c2045065a87f1fc141fabb004614f2c2cabb592e06c985b140892a3
+size 235987
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..59971c8ea5a8452bf64c94adf12f2e9101171891
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red car parked on a grassy area with trees in the background. The car is stationary and the focus is on the front right side of the vehicle, showcasing its shiny red paint and silver rims. The style of the video is a close-up shot, emphasizing the details of the car's design and the contrast between the red car and the natural surroundings. The car appears to be a modern model, with sleek lines and a shiny finish. The background is blurred, drawing attention to the car and making it the main subject of the video. The overall mood of the video is calm and serene, with the car parked in a peaceful, natural setting."
+    ],
+    "video_ids": [
+        "JXDXcXQG00Y_47_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red car, which is the main subject and matches the specified 'Object(s)' condition. The car is clearly visible and occupies the majority of the frame throughout the video, with no contradictions to the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a red car, specifically is the main object. clearly the description 'Object(s)' condition. The car is shown visible, occupies a majority of the frame, the video. fulfilling its significant to the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7f9e1d85674f23937bef87a617a550cf971a643f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:677ce789cbf58ece3034f3ee158a63e5d687a66cfdde8728570efba9e3eb40db
+size 149017
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..a7c2cbc0b9c2af3cd2bb39ad3e2f895edc9bb8c8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic scene at a race track. The first frame shows a car speeding down the track, its tires hugging the curve as it rounds a bend. The second frame shows the car continuing its journey, now further down the track, with the landscape stretching out behind it. The third frame shows the car approaching a chicane, a series of tight turns that require precise handling. The track itself is surrounded by a dry, arid landscape, with a few trees scattered here and there. The sky above is clear and blue, suggesting a perfect day for racing. The style of the video is a high-angle shot, providing a bird's-eye view of the action on the track. This perspective allows for a comprehensive view of the car's movement and the surrounding landscape."
+    ],
+    "video_ids": [
+        "EmD6pDS6nLk_7_0to194"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a car driving on a racetrack, fulfilling the 'Object(s): A car' condition. The car is visible throughout the sequence, navigating the track, and there are no elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a car driving on a racetrack, which the 'Object(s)' A car' condition. The presence is depicted in the frames, and the track, which there are no elements in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ea20812ef43b1054c806dfdc807bc3eb0aff342e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1303-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:504a891e9f1c5754f62d5741938ee5c48e48b5c1acfb1fafcc09bc1f107ac1e4
+size 267489
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..9423859f90969f1e29f1e05a6e247c0848c052b1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a single cupcake with a chocolate base, sitting on a metal surface. The cupcake is adorned with a colorful swirl of frosting in shades of pink, blue, and purple, creating a vibrant and playful appearance. The frosting is generously spread, with the colors blending together in a smooth and creamy texture. The cupcake is presented in a way that highlights its delicious and visually appealing nature, with the focus on the frosting and the contrast between the dark chocolate base and the brightly colored frosting. The metal surface provides a simple and uncluttered background, allowing the viewer to fully appreciate the cupcake's details and colors. The style of the video is straightforward and clear, with a focus on the cupcake and its decoration, without any additional elements or distractions."
+    ],
+    "video_ids": [
+        "lU4icAyfUE4_34_748to876"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A cupcake with a chocolate base and colorful frosting? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a cupcake in a black paper liner, which can be interpreted as a chocolate base, topped with colorful frosting in shades of pink, purple, and blue, along with chocolate swirls. This matches the core description of a cupcake with a chocolate base and colorful frosting.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a cupcake with a silver paper liner with which is be considered as a chocolate base. topped with a frosting in shades of pink and blue, and blue. which with a drs. The matches the description description of a cupcake with a chocolate base and colorful frosting.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..03c819c571748918e6d8a00985d9ea83044784b5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6cd5fcf7fd8b442c1ae48f6252f68ef322f69dbd251fadf2f6729bd98a07ef49
+size 108905
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..d42f5b51a375efc5e04301ade4a5b0d6531e8a75
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a dessert being eaten. The dessert is a slice of cheesecake, which is being scooped up with a spoon. The cheesecake is placed on a plate, and there are pieces of fruit, including strawberries and kiwi, arranged around it. The plate is on a table, and there is a bowl in the background. The style of the video is a simple, straightforward food video, focusing on the dessert and the process of eating it. The lighting is bright, and the colors are vibrant, highlighting the textures and colors of the food. The video does not contain any text or additional elements, and the focus is solely on the dessert and the act of eating it."
+    ],
+    "video_ids": [
+        "Ml5AAM_O6lE_46_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Cheesecake slice, spoon, strawberries, kiwi, plate, bowl? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a cheesecake slice, a spoon scooping from it, strawberries, kiwi, a plate, and a bowl in the background. All specified objects are present and identifiable in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showing showing a sliceecake slice, a spoon,oping the the, strawberries, kiwi slices and plate, and a bowl in the background. The the objects are present and identifiable in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ad72480514c50146a04475e91b4d02785bbbc20f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a155368b9708e33ffd26e2c978d0d4928253921ce9b11d20efceaedc1eed2d47
+size 142274
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..689641adf3093a82d47063a179956f1c8a7b35b6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of several pieces of chicken cooking on a grill. The chicken pieces are arranged closely together, with some showing signs of charring and browning, indicating they have been cooking for a while. Smoke rises from the grill, suggesting the heat is high and the chicken is being cooked thoroughly. The grill surface appears to be made of metal wire, allowing the heat to circulate evenly around the chicken. The background is slightly blurred, keeping the focus on the chicken and the grill. The lighting is natural, likely from an outdoor setting, enhancing the golden-brown color of the chicken. There are no visible characters or animals in the scene, and the camera remains stationary throughout the video, providing a steady view of the cooking process. The video emphasizes the texture and color of the chicken, highlighting the appealing appearance of the dish as it cooks."
+    ],
+    "video_ids": [
+        "c7cbfd7d4e25e62378ec12f1f18a3f10cf85e7af70b4ada58b4ec0effe0871de"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Several pieces of chicken? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows multiple pieces of chicken, specifically chicken legs, being grilled over coals. The description of 'Several pieces of chicken' is accurately fulfilled by the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows several pieces of food on which chicken wings, which grilled. aals. The presence of 'Several pieces of chicken' is fulfilled represented by the visual content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..786d9c643ed488eb4b66d22f6ca263edd24eee11
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:31a8739b501160e651dec7889b9be0d4ddf12904d986d57efc1a76188a900a76
+size 89285
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..eb65de01809e6c5d2b94bb2e0e0f525b0c5c060e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a robot dressed in a white space suit with a gold helmet, standing in a space-themed environment. The robot is adorned with NASA and GM logos on its chest, indicating a partnership between the two organizations. The robot is positioned in front of a blue wall, which adds a sense of depth to the scene. The overall style of the video is futuristic and sleek, with a focus on the robot as the main subject. The lighting is bright and even, highlighting the details of the robot's suit and helmet. The video does not contain any text or additional objects, and the robot remains stationary throughout the frames. The video is likely intended to showcase the capabilities of the robot and the collaboration between NASA and GM."
+    ],
+    "video_ids": [
+        "llix751qov0_15_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A robot dressed in a white space suit with a gold helmet, adorned with NASA and GM logos.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a robot wearing a white space suit with a gold helmet, and both the NASA and GM logos are clearly visible on its chest. The description matches the visual content without contradiction, even though the background includes additional elements like machinery and equipment, which do not conflict with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a robot dressed a white space suit with a gold helmet, which it NASA NASA and GM logos are visible visible on the chest. The background align the visual elements of any.\"\n and though the background appears a elements like a and a, which are not conflict with the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0a11fde1a9de565e8fb939fd5c858a211a3ce3fe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f1082f3e5b7593b8dfcea81357db1af5e35c9e4222a459ef84801b82e260548e
+size 153096
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..005fdbb8d24d74d7939a89540b8626a4170467d1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with a furry hood covering her head, looking down with a contemplative expression. The fur is light brown and appears soft. The woman is indoors, with a plain white wall in the background. There are no other people or objects in the scene. The lighting is soft and even, suggesting an indoor setting with ambient lighting. The style of the video is simple and straightforward, focusing on the woman and her expression."
+    ],
+    "video_ids": [
+        "5NSSGk2CwTI_29_0to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with a light brown, soft furry hood covering her head, looking down with a contemplative expression.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a light brown, soft furry hood that covers her head. She is looking down with a contemplative expression, which aligns with the description. The presence of hands adjusting the hood and the background elements (hanging caps) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a woman wearing a light brown, fluffy furry hood that covers her head. She is looking down with a contemplative expression, which aligns with the description provided The fur of the in the fur does the slight elements dolikeanging clothes and do not contradict the core description and}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..29a4784df4d986c4549b898335bc27c3d0276c93
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4630ae963527ef566cd0903265f8033e051bbce756a695ce60ad0cc6f2a89363
+size 155142
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..93f0c9a58c2612ef31db4e519b8ec465e1f1154a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men sitting at a table, engaged in a conversation. The man on the left is wearing a black cap and a black shirt, while the man on the right is wearing a white shirt. They are both sitting with their hands clasped on the table. The table is wooden and there is a red light shining on it. The background features a colorful wall with various decorations and a window. The overall style of the video is casual and relaxed, with a focus on the interaction between the two men."
+    ],
+    "video_ids": [
+        "1DB2O2ZhsPo_3_110to271"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men sitting at a table, one wearing a black cap and shirt, the other wearing a white shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men sitting at a table. One man is wearing a black cap and black shirt, and the other is wearing a white shirt. These details match the description exactly, and no elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men sitting at a table. One man is wearing a black cap and a shirt, while the other is wearing a white shirt. The details match the description provided, indicating there additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2f7f6a090ee4a029358459b64320dd15b8d3fa8c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d93eb27a8f0677caad4cf3fd58c0ea50a6f6e6b295b4b67f6cdc3029bc6530e1
+size 94985
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..956f0a95c8de7c9c9c73cf017dd8079fdfd94d0e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man's journey on an airplane. He is seen in a black t-shirt and a backwards cap, walking down the aisle of the airplane. The airplane is filled with rows of seats, all neatly arranged. The man's journey takes him past a flight attendant, who is standing at the front of the plane. The video is a snapshot of a moment in time, capturing the anticipation and excitement of travel."
+    ],
+    "video_ids": [
+        "WlsYx3sZoVU_111_54to201"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black t-shirt and backwards cap, a flight attendant? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a black t-shirt and a backwards cap, who is the main subject. A flight attendant, identifiable by her uniform and actions (walking through the aisle), is also visible in the background. Both elements described in the condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a man in a black t-shirt and a backwards cap, and appears walking focus subject. Additionally flight attendant is dressed by their uniform and cap,standing down the aisle), is also present. the background. The elements match in the question are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..951ed060a4010078b57e5445d8cff2de61b1595a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:775f0d3680f3be36483f803d5f7b0a55511dd017bef04bcb19ac244d7e69c13d
+size 161798
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..56b0c8c16a0c6fe755d2165af764d5d8951b9835
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a bowl of food, which appears to be a type of Asian cuisine, possibly beef and vegetable stir-fry. The bowl is filled with slices of beef, shredded vegetables, and a single orange slice. The food is garnished with a sprig of green onion. The beef is cooked to a medium rare, and the vegetables are bright and fresh. The orange slice adds a pop of color to the dish. The food is served on a white plate, which contrasts nicely with the vibrant colors of the food. The overall style of the video is simple and straightforward, focusing on the food and its presentation."
+    ],
+    "video_ids": [
+        "5MVHFHyZQW0_13_0to172"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bowl of food, slices of beef, shredded vegetables, orange slice, green onion sprig? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a bowl of food topped with slices of beef, shredded vegetables (including bean sprouts and green onions), an orange slice (which appears to be a tomato or similar round red vegetable), and a green onion sprig. All elements described in the condition are present in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a bowl of food that with slices of beef, shredded vegetables,which carrots sprouts and carrots onions), an orange slice,likely appears to be a slice slice a fruit fruit fruit), and a green onion sprig. The elements match in the condition are present in the image.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..182991a0ee6dc0217e41d3fdafee17b3475d9bff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1304-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8649cf6779e6cabf1673115f3454da40eab91d2c9c913e70bbbd7fddb31c3f77
+size 126539
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..6fd95a152013f2fc2e14084145f91f986e08f18b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video begins with a dark screen featuring the text \"Thanks to\" prominently displayed in white at the center. Below this text, four logos are arranged horizontally. From left to right, the first logo is for Cambridge University Press, featuring a shield emblem with a red cross and a red and white checkered pattern. The second logo represents ANU (The Australian National University), showcasing a blue and gold shield with a crown above it. The third logo is for Questacon, the National Science and Technology Centre, depicted with a stylized orange and black design. The fourth logo is for the Australian Government, showing a white kangaroo and emu on a green background. The scene remains static throughout the video, with no visible movement or changes occurring."
+    ],
+    "video_ids": [
+        "ea5f0678587535c550a79ed7a9c381a862fe5b858acb7310fb4de23073251efc"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Text 'Thanks to', logos for Cambridge University Press, ANU, Questacon, and Australian Government.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully displays the text 'Thanks to' centered on the screen, along with the logos for Cambridge University Press, ANU, Questacon, and the Australian Government, as specified in the condition. All elements are clearly visible and correctly positioned.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video contains ful logos logos 'Tr to' at at the screen, followed with logos logos for Cambridge University Press, ANU, Questacon, and Australian Australian Government. which described in the condition. The the are present visible and match positioned,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..eb70d53684eb02975984070e044bdacc3fbd756b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4155fd96cf67189651a5d5c34dcaf086f4b87f174659c6954bbfa44dac6386a9
+size 99832
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..1a31e84b13cdb150fd70e4eb6664e8b061bb00de
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a makeup room, preparing her look. She is wearing a black shirt and has her hair styled in an updo. Her makeup is dramatic, featuring green eyeshadow and a green headpiece. She is adjusting her headpiece in front of a mirror. The room has a white shelf in the background, which holds various makeup products. The overall style of the video is a behind-the-scenes look at the process of getting ready, focusing on the woman's transformation through makeup and styling."
+    ],
+    "video_ids": [
+        "cDzJoC9JTcI_25_47to223"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a mirror, a black shirt, an updo hairstyle, a green headpiece, and green eyeshadow.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman wearing a black shirt, with an updo hairstyle and a green headpiece (a nest with eggs). She also has green eyeshadow. A mirror is visible to her left. All core elements described are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a woman wearing a black shirt, with an updo hairstyle and a green headpiece.likely green-like a). She is has green eyeshadow applied The mirror is visible in the right, The the elements of in present in accurately depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..861f1a9d11c2ab4f774a689f623b27077852d4bc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf14df9bb87fb1169cf41c5fce50b901ab1b626f4cc46589e8bd7fc9c49017ec
+size 206719
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..9299e49bf17fa0e2344241743832d3f978a3993e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a plate of fries with parsley garnish. The fries are golden brown and appear crispy. The parsley garnish is bright green and sprinkled on top of the fries. The plate is white and the fries are piled high on it. The background is blurred, but it appears to be a kitchen or dining area. The style of the video is a simple, straightforward food shot with a focus on the fries and their garnish. The lighting is bright and even, highlighting the colors of the fries and parsley. The video does not contain any text or additional elements."
+    ],
+    "video_ids": [
+        "NjK2jHPl1jM_91_104to240"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Fries, parsley garnish? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a dish of fries topped with a garnish of chopped parsley. The fries are golden and appear crispy, and the parsley is finely chopped and evenly distributed over the top, matching the description. No conflicting elements are present that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a plate of fries garn with a garnish of parsley parsley. The fries are golden and appear to, and the parsley is fresh chopped and placed distributed on the top, fulfilling the description of There additional elements are present.\"\n would the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a5dbdaed07be41e8b60f50ebcf85935ef8a241d1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c43dcba624a4dc700c4d37267e59b8bcd819f7b06a5833d81657aba702584ead
+size 119463
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..0dbcd82f10287e5dbfaba49b407416789ab2c94b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with long brown hair, wearing a blue dress, and red lipstick. She is speaking and looking to her right. The background is dark, which highlights her as the main subject. The style of the video is a close-up shot, focusing on the woman's face and upper body. The lighting is soft, creating a gentle glow on her face. The video captures a moment of her speaking, possibly in a public setting or event."
+    ],
+    "video_ids": [
+        "HK5aiZGu8L0_4_97to278"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with long brown hair, wearing a blue dress, and red lipstick.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with long brown hair, wearing a blue dress, and red lipstick, which matches the description. The background is black, and there are no conflicting elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with long brown hair, wearing a blue dress, and red lipstick. which matches the description provided The background appears not and but the are no additional elements in contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c80727e357b2243c2003e68ec57e3aa1137cc52b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ca1768fcbdf301346b25610d37a8f6217a7e3a18a0c96119b05781f3d4c8dd42
+size 117790
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..b1ba983d659efdb8670f961f925bfe578fdda78d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a cowboy hat and glasses, standing in a courtyard with a stone building in the background. He is wearing a plaid shirt and a blue bandana around his neck. The man is looking directly at the camera, and his expression is serious. The lighting in the scene is natural, suggesting it is daytime. The man's attire and the setting give the impression of a western or cowboy theme."
+    ],
+    "video_ids": [
+        "L61r_nMWp6M_33_0to117"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a cowboy hat and glasses, wearing a plaid shirt and a blue bandana around his neck.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a cowboy hat, glasses, a plaid shirt, and a blue bandana around his neck, which matches the description. The background and additional elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a cowboy hat, glasses, a plaid shirt, and a blue bandana around his neck. which matches the description provided The background appears additional elements do not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a383198c42c9736ca4221d8322984da1c3e7d5c8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c31d0a71a59bec877745e47e3688f3fd6e3722016139126eb95a065534965585
+size 240794
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..d85e398a6b5c631d82546d0173c7786a18dc6966
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a baseball game. The main focus is a baseball player, dressed in a vibrant red and white uniform, who is in the midst of swinging his bat. The player's stance and the motion of the bat suggest a powerful swing, possibly aiming for a home run. The player's uniform is clearly marked with the number 34, indicating his identity within the team.  In the background, a crowd of spectators can be seen, their attention riveted on the unfolding action. The spectators are a mix of individuals, all engrossed in the game, adding to the atmosphere of anticipation and excitement.  The setting appears to be a professional baseball stadium, characterized by its well-maintained field and the presence of a large crowd. The stadium's architecture and the crowd's attire suggest a sunny day, perfect for an outdoor sporting event.  Overall, the video is a snapshot of a thrilling moment in a baseball game, capturing the player's skill, the crowd's enthusiasm, and the vibrant atmosphere of the stadium."
+    ],
+    "video_ids": [
+        "rVKHolhrRIM_5_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A baseball player in a vibrant red and white uniform with the number 34, a baseball bat being swung.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a baseball player wearing a vibrant red and white uniform with the number 34, and he is in the act of swinging a baseball bat. The player's posture and equipment are consistent with the description, and there are no elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a baseball player in a vibrant red and white uniform with the number 34, and the is swinging the process of swinging a baseball bat. The player's attire and the align consistent with the description, and the are no additional in contradict the given description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fe247d83107a73d2dccacfafe9c881e348496609
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:276f1a21d0485770162b8332f0b1c63e2e1de1c5aa3d78440402e5af5457c212
+size 234734
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..3142baf0fa909150ee0b5ea08011453fd20cdb61
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the transformation of a large construction site over time. In the first frame, the site is a vast expanse of dirt and rubble, with several cranes and construction equipment scattered across the landscape. The second frame shows the site beginning to take shape, with the foundation of a large building visible and more construction equipment in place. By the third frame, the building has risen significantly, with its structure clearly visible against the backdrop of the city skyline. The cranes and construction equipment are still present, but the site is now bustling with activity as the building nears completion. The video is a testament to the power of human ingenuity and the marvels of modern construction."
+    ],
+    "video_ids": [
+        "hElaNvauzL8_1_0to194"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Cranes, construction equipment, a large building? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows multiple cranes and various construction equipment scattered across a large construction site. A large building is in the process of being constructed, with visible structural elements and scaffolding. These elements align with the 'Object(s)' condition specified.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a cranes and construction construction equipment at across a construction construction site. As large building is also the process of being constructed, as the progress elements and aolding. The elements align with the 'Object(s)' condition specified in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0572fa479b07f8427ec60330f0f8479f0ab2d255
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2718a5c8382871df55939dea79336339f85d284fc2d776cb534f114d26998967
+size 343703
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..4539afd780700958a1aaad75f65331e3de716492
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a split-screen montage of two men and a bustling city scene. On the left, a man in a gray hoodie and red sunglasses gives a thumbs-up, while on the right, another man in a black jacket and short beard looks on. The city scene in the background shows a busy street with people walking, buildings, and a traffic light. The style of the video is a collage of urban life and personal interaction, capturing the essence of a city's hustle and bustle."
+    ],
+    "video_ids": [
+        "MafaFGmbxM8_19_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men (one in a gray hoodie and red sunglasses giving a thumbs-up, one in a black jacket and short beard looking on).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features two men as described: one wearing a gray hoodie and red sunglasses who gives a thumbs-up gesture, and another in a dark hoodie with a short beard who looks on. The presence of a split-screen with a crowd does not contradict the core description of the two men.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men standing described: one in a gray hoodie and red sunglasses giving is a thumbs-up,, and the in a black jacket with a short beard who looks on. The setting of a traffic-second effect both blurred in not contradict the core description but the two men and}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ef4eddaf1c9766c369aa5e5bf28f58cd63f4241e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1305-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b8dd5949c00c2cccdbe895ad646738ede5d4f25efa1d3d64c1fd50b5d493c42c
+size 192196
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..f967a15efef67fb620ce2b8f6f9ee7f97950ef4a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a brown SUV parked in a parking lot. The car is stationary and there are no people or other vehicles in the vicinity. The car is parked in a designated parking space, and the ground is wet, suggesting that it has recently rained. The car is the only object in the frame, and there are no other objects or actions taking place. The style of the video is straightforward and unembellished, focusing solely on the car and its immediate surroundings."
+    ],
+    "video_ids": [
+        "RC9z37uU9AQ_8_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A brown SUV parked stationary.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a brown SUV parked stationary in a parking lot, with no movement or changes in position throughout the frames. The vehicle remains consistently in the same spot, matching the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a brown SUV parked in in a parking lot. which no additional or additional in the. the frames. The environment is in in the same spot, fulfilling the description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5ea5b85f761a21d8e713a92f1a59ce870f662269
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:afb3ca8fbb2a724afcd82c39bfb71b36f68e57461c73948ce26513d4256358d4
+size 176298
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..ed7463351c7d2810409482713258aa425ca2fdaf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a formal meeting taking place in a conference room. The room is adorned with a large wooden table, around which five individuals are seated. The participants are dressed in business attire, with some of them wearing ties. The table is cluttered with various documents and papers, indicating an ongoing discussion or negotiation. The room is decorated with a large flag, adding a touch of national pride to the setting. The overall atmosphere suggests a serious and professional environment."
+    ],
+    "video_ids": [
+        "CX5YvKMqWNM_48_24to205"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Five individuals seated around a large wooden table. Various documents and papers are on the table.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows five individuals seated around a large wooden table, with various documents and papers spread across the table. The scene is consistent with the described condition, and no elements contradict it.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows five individuals seated around a large wooden table. with various documents and papers spread across the table. The setting is set with the description ', and there additional contradict the.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9e547da265051d965f52d4d7693997198ad48012
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e1a48455aca64e18a4ca84caf9c256bbdb98ab9e4b02d93ab43e3aabf8004edb
+size 212829
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..d2112605c24fd915926f7f9765294f40b09aebbb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man sitting in front of a microphone, likely in a home studio or gaming setup. He is wearing a black t-shirt and has a beard. The room is dimly lit with blue and purple lights, creating a moody atmosphere. On the wall behind him, there are shelves filled with various figurines and collectibles, suggesting that he might be a gamer or a collector. The man appears to be speaking into the microphone, possibly recording a podcast or a live stream. The overall style of the video is casual and informal, with a focus on the man and his surroundings."
+    ],
+    "video_ids": [
+        "1u7ljmSd55Y_5_133to308"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a microphone, various figurines and collectibles on the wall.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man sitting in front of a microphone, with various figurines and collectibles visible on shelves in the background. These elements match the 'Object(s)' condition described, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man speaking in front of a microphone, which various figurines and collectibles displayed on the in the background. The elements match the descriptionObject(s)' condition described in indicating there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cbd42436195c22564a5287904ca69e57e377335f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d9b165042e0fd74f8f85efc646d107e86d3b78a5c7f9162432e7cefa7036c4d2
+size 145115
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..5c0f06674e9e6324c47432217bbdc23b88bf339b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man wearing a white face mask, standing in a park-like setting with trees and a paved path. The man is looking directly at the camera, and his expression is neutral. The style of the video is a straightforward, real-life depiction with no special effects or artistic filters. The focus is on the man and his mask, suggesting a theme related to health or safety precautions. The background is natural and serene, providing a contrast to the man's presence."
+    ],
+    "video_ids": [
+        "EJ43ek3B0A8_1_0to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man wearing a white face mask, standing and looking directly at the camera with a neutral expression.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a white face mask, standing outdoors and looking directly at the camera. His expression appears neutral, and the background elements (trees, buildings, park features) do not contradict the core description. The focus remains on the man as requested.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a white face mask, standing outdoors, looking directly at the camera. His expression is neutral, and the setting suggests,trees and green, and-like) do not contradict the description description. The video is on the man, the.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5fd9017dfa196e878673a36966cd9cf6f66bb300
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60f3a867b850b30edfd4fab59b33e13d4dba0f28e0a2b3668ba6c6f12cf92966
+size 89684
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..6a24bf5480151641e0ed644d2ce54be290fad073
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman wearing glasses is seated at a table in a restaurant. She is wearing a white sweater and is holding a small piece of food in her hand, examining it closely. On the table in front of her, there is a bowl of salad and a cup of coffee. The restaurant has a warm and inviting atmosphere, with a wooden table and a comfortable chair. The woman appears to be enjoying her meal and the ambiance of the restaurant."
+    ],
+    "video_ids": [
+        "Q3p-Wyc6FM8_18_479to627"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a bowl of salad, a cup of coffee? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman sitting at a table, eating with a fork. A bowl of salad is visible on the tray next to her, and a cup (likely coffee or tea) is also present on the tray. All core objects mentioned in the condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a woman sitting at a table. holding a a small. There bowl of salad is visible on the table in to her, and a cup ofwhich containing) tea) is also present. the table. The elements elements ( in the condition are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..764f5595609184bf1e4dfa0d3a5193212928336a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e7ced0a9ae19c0b57955ff1af720372f5f75d103b37c169b706a31ffb23c892
+size 111052
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..03d82dc485a892daf87b8ac23f06c252f9e1080b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young woman with blonde hair, wearing a white t-shirt and a black microphone clipped to her collar. She is standing against a plain white background. The woman appears to be speaking or presenting, as suggested by the microphone. The style of the video is simple and straightforward, with a focus on the woman and her message. The plain background ensures that the viewer's attention is drawn to her. The video is likely intended for an audience interested in listening to the woman's speech or presentation."
+    ],
+    "video_ids": [
+        "AytwvGbqcJI_13_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: young woman with blonde hair, white t-shirt, black microphone clipped to her collar? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman with blonde hair wearing a white t-shirt, and a black microphone is clearly clipped to her collar. These elements match the description exactly.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman with blonde hair wearing a white t-shirt. and there black microphone is clipped clipped to her collar. The elements match the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1d1d134ae3ed842366a1436687dfdd09b27e3584
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d4917941b985e44c30d7e417d86e7b2dbbdca7685a1bb8d4696314ef67dd7d05
+size 106197
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..3e4250a3ce9646680a75d2c810e6f98f17400473
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up of a futuristic helmet with a green visor. The helmet is predominantly white with yellow accents. The visor is green and covers the entire face of the helmet. The helmet has multiple buttons and switches on the side. The background is blurred, but it appears to be a rocky cave or tunnel. The style of the video is realistic with a focus on the helmet, suggesting it might be from a video game or a science fiction movie. The lighting is dim, highlighting the helmet's details."
+    ],
+    "video_ids": [
+        "J5DjrBrF_ao_15_0to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Futuristic helmet with green visor, multiple buttons and switches? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a futuristic helmet with a green visor, and details such as multiple buttons and switches are visible on its surface. The helmet's design, including its metallic structure and color scheme, aligns with the description. No conflicting elements are present that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a futuristic helmet with a green visor, which it such as multiple buttons and switches are clearly on the front. The helmet's design includes including the color and and the scheme, aligns with the description of The conflicting elements are present, would the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6b9d514e1940d32b898cfb5a503b6ba7577e1778
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b08a6787aef87b5b3d091d37877dfe823f97c219f730cfa639c796c8f5deb4b
+size 189558
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..9eaf783a8afb7ced395e69fe0eccecf7e2f66e7e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up shot of a bird in flight, showcasing its intricate details and dynamic movement. The bird, likely a pigeon or dove, is seen from below, emphasizing its wings and tail feathers. Its plumage displays a mix of white, gray, and brown hues, with the wings spread wide, revealing the texture and pattern of the feathers. The bird's head is turned slightly to the side, giving a glimpse of its beak and eye, which appear alert and focused. As the bird soars through the air, its wings flap rhythmically, creating a sense of motion and grace. The background is blurred, suggesting a natural outdoor setting with soft, diffused lighting that enhances the bird's features. The BBC logo in the top left corner indicates the source of the footage, while the number \"8\" in the top right corner suggests it might be part of a series or episode. The overall scene conveys a serene and majestic moment in nature, highlighting the beauty and agility"
+    ],
+    "video_ids": [
+        "fe400e488bcfb8a746c7847813ca125401d1fa90fa393187d41914a9921c9fe1"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bird (likely a pigeon or dove)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bird in flight with the characteristic features of a pigeon or dove, including its body shape, wing structure, and coloration. The bird is clearly visible and matches the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bird in flight, features characteristics features of a pigeon or dove, such its plum shape, feather structure, and plumation. The presence's the the and the the description of.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e8ec82fe9ae0c48c9366e72c98634881c6dd0a6f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1306-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50194b10a4ea0034399e823873a207b5be62719cac8a730dfad2eee0d9b05291
+size 256019
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..58c27f94dc7fb9e682c41e2d1a4c281dd71df4b8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a cartoon character with a blue and white face, wearing a black and orange outfit, is sitting at a wooden table. The character is holding a plate of green food, possibly salad, and there is a bottle of ketchup next to the plate. The character appears to be enjoying the meal. The background features a green cabinet with a red and white object inside, and a wooden door. The overall style of the video is cartoonish and colorful, with a focus on the character and the meal."
+    ],
+    "video_ids": [
+        "tNJvUmCtLWk_244_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A cartoon character, a plate of green food, and a bottle of ketchup.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a cartoon character (a vampire-like figure) sitting at a table with a plate of green food (appearing to be spinach or similar greens) and a bottle of ketchup (red bottle with tomato imagery). All core objects mentioned in the condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a cartoon character,a blue-like figure with sitting at a table with a plate of green food inwhicharing to be a or a leaf) in a bottle of ketchup.with with with a design). The the elements are in the condition are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0816423d29a32eb3fc866286f61b9858db6a3bf5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a79430a5291ac6ccaf214aff3fd084102c3f8dffe740b2069232dfeb28123c36
+size 137664
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..db6e72e85112701888205cdf1005495cce59a8e4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animation that showcases a colorful and vibrant depiction of a molecular structure. The animation features a variety of molecules, each represented by a different color and shape, interconnected by green bonds. These molecules are set against a backdrop of blue bubbles, which could represent cells or a cellular environment. The style of the animation is scientific and educational, aiming to illustrate the complex structure of molecules and their interactions within cells. The video is likely used for educational purposes, such as teaching chemistry or biology."
+    ],
+    "video_ids": [
+        "ykyq54ZSgGI_3_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Various colored and shaped molecules interconnected by green bonds.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video displays various colored molecules (green, red, blue, yellow) that are interconnected, and green bonds are clearly visible connecting these molecules. The core description is accurately represented, with no significant contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a colored and withpink, red, pink, and, interconnected are interconnected by with the bonds are visible visible, these molecules. The molecules elements of well represented in and no significant contradictions or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ec00db3afa5e989d5a8008ecda2a494d24bfc2bb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6038959c93a3a9e3e8ab327b29b3264f5d3385880471f3ba281e8173b8511461
+size 264547
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..513a738c1331ab9833065dc67df6b44944860c41
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two young men in a park-like setting, engaged in a conversation. The man on the left is wearing a dark-colored shirt and glasses, while the man on the right is wearing a bright yellow hoodie with the word \"ARY\" on it. They are both looking down at something in their hands, possibly a phone or a piece of paper. The background is blurred, but it appears to be a sunny day with trees and greenery. The style of the video is casual and candid, capturing a moment of interaction between the two men."
+    ],
+    "video_ids": [
+        "BHWahwyH_TU_39_27to173"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two young men? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two young men sitting outdoors, engaged in conversation. Both are visible and identifiable as the primary subjects, fulfilling the 'Object(s): Two young men' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two young men. together, one with looking. One are wearing and the as young main subjects of fulfilling the 'Object(s)' Two young men' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..aff37ee959aade91fe228e24d98a41923cf420f2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:241aecbdfe521f4790a7b157d56ecf188be36ae71a47c01e0a9e138c17cfa7f6
+size 141827
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..f00bee46c31f916a619e52652b1cbadbdb69f86d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a chef in a professional kitchen, preparing a dish. The chef is wearing a white chef's coat and is focused on stirring a pot of food on a stove. The stove is a modern, stainless steel model with a digital display and control panel. The chef is using a long-handled spoon to stir the contents of the pot. On the counter next to the stove, there are several bowls and cups, suggesting that the chef is in the process of preparing multiple ingredients for the dish. The kitchen is well-lit and clean, with a professional and organized appearance. The style of the video is a straightforward, documentary-style recording of the chef's actions, with no additional embellishments or artistic effects."
+    ],
+    "video_ids": [
+        "IGxlQfyjK2Y_3_24to182"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chef, stove, long-handled spoon, bowls, and cups.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chef in a white uniform using a stove with a pot, stirring with a whisk (a type of long-handled spoon), and there are bowls and cups on the counter. All the specified objects are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person in a professional uniform cooking a long to a long and which with a long,which long of long-handled spoon), and there are bowls and cups visible the counter. The the elements objects are present and the with the scene.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2e9a7d512edcc82596288806ab4741bdb84b7fb6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea94e2272ed5735371f00a4a14900bebb1160e76409466178a22229d49a2cd1c
+size 121027
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..690bd68d0e816b13a3bf582dcee0687d7c620db1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic scene on a soccer field. Two soccer players, both dressed in blue uniforms, are in motion. The first player is seen running towards the right side of the frame, his body leaning forward in anticipation. The second player is seen running towards the left side of the frame, his arms outstretched as if he's about to make a move. The background reveals a well-maintained soccer field with a goal post visible in the distance. The players' movements and the setting suggest an intense game in progress."
+    ],
+    "video_ids": [
+        "UpwAyCDzGtw_24_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two soccer players in blue uniforms? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two soccer players wearing blue uniforms, which matches the 'Object(s)' condition. Although there are additional elements like poles and a background banner, they do not contradict the core description of the two players in blue uniforms.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two individuals players wearing blue uniforms, which matches the descriptionObject(s)' condition. The the are additional elements like the and a goal that, they do not conflict the core description of the two soccer in blue uniforms.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e135edb9b382530170a1505e31159ed3e33c7cab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:516b8874c0229cb295fdfb3cab556571ae91756cf35ab6fea69cd26a5f9d8b24
+size 403097
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..29d1b532e55149e9a8736eaed32650be0d25a9c4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a cyclist riding down a winding mountain road. The cyclist, dressed in a red jersey and a white helmet, is seen pedaling along the road, which is lined with a guardrail on one side and a steep drop-off on the other. The road itself is a two-lane highway, with white lines marking the lanes. The cyclist is riding on the right side of the road, following the rules of the road.  The backdrop of the video is a stunning mountain landscape. The mountains rise steeply on both sides of the road, their rugged terrain covered in a mix of green and brown vegetation. The sky above is a clear blue, suggesting a sunny day. The cyclist's journey down the mountain road is set against this breathtaking natural scenery, creating a sense of adventure and freedom. The video is likely shot from a car following the cyclist, capturing the cyclist's progress down the road and the stunning landscape around them."
+    ],
+    "video_ids": [
+        "FUrodNN5Aag_9_23to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A cyclist in a red jersey and white helmet, a guardrail, a steep drop-off, a two-lane highway with white lines, and the mountain landscape.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a cyclist in a red jersey and white helmet riding on a two-lane highway with white lines. The road is bordered by a guardrail, and there is a steep drop-off visible on the side. The background features a rugged mountain landscape, matching all specified elements in the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a cyclist in a red jersey and white helmet riding on a two-lane highway with white lines. The presence is bordered by a guardrail, and the is a steep drop-off visible on the right. The background features a mountain mountain landscape, which the the elements.\"\n the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2a36cbde359713079be7fb970af0b218507e3880
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8076709c4daba52ba1752d7c4f4b1a560d599ef77ae2fc6cf8f720bede96a480
+size 252087
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..f49bd804727ecb1190403530c368ca2ee24f5540
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man wearing a black cap with neon green accents and a gray hoodie. He is smiling and looking to his left. The background consists of a black wall with a large, white, cursive \"S\" on it. The lighting in the scene is bright, with the \"S\" illuminated, creating a contrast between the dark background and the brightly lit letter. The man appears to be in a good mood, possibly enjoying a moment of relaxation or amusement. The overall style of the video is casual and informal, with a focus on the man's facial expression and the brightly lit background."
+    ],
+    "video_ids": [
+        "BFiJEYxdmUg_8_0to126"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man wearing a black cap with neon green accents and a gray hoodie, smiling and looking to his left.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black cap with neon green accents and a gray hoodie, and he is smiling while looking to his left. The description matches the visual content without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black cap with neon green accents and a gray hoodie. smiling he is indeed while looking to his left. The description matches the visual elements of any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..339d82e50a171b7a5e1523966b76a31c3b9686bf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a902c3d104eb432a1b2318dd7190409d632fab94c95584949bbd9c48d02ad65e
+size 177716
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..a2dd0acb97d28a8f78141bf7778d79c1dafeb36c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up of a black and green motorcycle helmet with a clear visor. The helmet is shown from the side, with the visor partially open, revealing the interior of the helmet. The helmet has a reflective surface, and the visor has a slight tint. The helmet is positioned against a blurred background, which suggests that the focus is on the helmet itself. The style of the video is a product demonstration, showcasing the features of the helmet. The video is likely intended for a motorcycle enthusiast audience, highlighting the design and functionality of the helmet."
+    ],
+    "video_ids": [
+        "B6BbQ0-2TYQ_13_0to175"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black and green motorcycle helmet with a clear visor, partially open to reveal the interior.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black and green motorcycle helmet with a clear visor, and the visor is partially open, revealing the interior. The description matches the visual content accurately, with no contradictory elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close and green motorcycle helmet with a clear visor. which the visor is partially open, revealing the interior of The helmet matches the visual content of, with no additional elements present}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f3f80e5f3421f590db2191944fc7aae02e365155
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1307-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c69ef01d23fd7665c60bb5690a00eae2390c4a19ebd3870943e308f50ae2f167
+size 135903
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..b4a094302d3eb344365d86e4ff05cfce7758fe11
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing on a soccer field, with mountains in the background. He is wearing a black jacket with white stripes and has a beard. In the first frame, he is looking down, in the second frame, he is looking up, and in the third frame, he is looking to the side. His hands are raised in front of him, and he appears to be speaking. The style of the video is casual and candid, capturing a moment in the man's day."
+    ],
+    "video_ids": [
+        "YIVvOdFi8wk_0_0to175"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, mountains, soccer field? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a man (the main subject), mountains in the background, and a soccer field as the setting. The man is visible throughout the video, the mountains are in the distant background, and the soccer field is the ground surface where the man is standing and lying down. Additional elements like text overlays or camera effects do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features fulfills the 'Object(s)' condition as featuring showing a man,likely main subject), mountains in the background, and a soccer field. the setting. The man is consistently in the video, and mountains are consistently the background background, and the soccer field is visible foreground on on the man is standing and gest down. The elements like the or or other angles do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3c982f68ec468ad26d618cb8eb622ac5bb904af3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a839709dda88cd10a0b9f254befc9ec89c94e13016d413c49191ec49cd75f4d
+size 241702
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..759346c3c35797d86a2e86c3479a8d215842f762
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young man in a white t-shirt standing in a living room. He has short hair and is looking directly at the camera with a slight smile. The living room has a fireplace with a mantel, and there are framed pictures on the wall. The lighting in the room is soft and warm, creating a cozy atmosphere. The man appears to be in a relaxed and comfortable setting."
+    ],
+    "video_ids": [
+        "Uzs5RYJZBkU_81_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Young man in a white t-shirt with short hair, looking directly at the camera with a slight smile.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man wearing a white t-shirt, with short hair, looking directly at the camera. His facial expression includes a slight smile, which aligns with the description. The background elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man wearing a white t-shirt and which short hair, and directly at the camera with He expression expression is a slight smile, which aligns with the description. The background,, not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a9802cc7ba6639f22f87a26844a621c21b5ab721
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75f1c033e5dd436281cfb21e3c69bc05f4d3203ba05ca8b95f21acf953d5eee0
+size 70312
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..d580a6d8c9ed6464cf66cb8c58b107a27834a2d0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows an elderly man with long gray hair, wearing a blue t-shirt with a logo on it, sitting in a booth with a brown leather seat. He is looking directly at the camera with a slight smile on his face. The background features a window that offers a view of a city street with buildings and shops. The lighting in the room is bright, suggesting it's daytime. The man appears to be in a relaxed and comfortable setting, possibly a cafe or restaurant. The style of the video is candid and informal, capturing a moment of the man's day in a natural and unposed manner."
+    ],
+    "video_ids": [
+        "DRMgKyn94GE_6_0to155"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An elderly man with long gray hair, wearing a blue t-shirt with a logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an elderly man with long gray hair, wearing a blue t-shirt with a visible logo (Leicester City). The core description is accurately represented, and background elements do not contradict this.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an elderly man with long gray hair, wearing a blue t-shirt with a logo logo.thoughvi).). The description elements of largely represented in and there elements do not contradict the.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..04fd2a9d118cda28f59a48bd3baab78fd42bcdd1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:05b0c3823fa9ed6fba12d003280f9dc31c94112be07f2faf670686709cb138ea
+size 231017
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..d4bb8acd1706b7b888048c36600e61d41e931130
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man standing inside a car, leaning over the steering wheel with his hands on the wheel. He is wearing a black jacket and a blue hat. The car is a modern, black SUV with a sleek design. The interior of the car is well-lit, and the man appears to be in a good mood, smiling as he looks out the window. The car is parked in a lot with a clear sky and a mountain in the background. The style of the video is candid and informal, capturing a moment of the man's day."
+    ],
+    "video_ids": [
+        "KK2vr8zFFDM_31_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a black jacket, a blue hat, a steering wheel, a window? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a black jacket and a blue hat, leaning into the car with his hands on the steering wheel and the car window open. These elements match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing a black jacket and a blue hat, sitting on the steering's his hand on the steering wheel, looking window window partially. The elements match the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1b4562e589c14ea23e263839d1b85c1be3451a06
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03815bc7677868ac733dbe679846e882cc6cf2d676f4b977f3e6607a2621e45f
+size 103486
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..f806d85e87c4807df044bbb13ea4cf645fc8855e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, three astronauts are seen in a space station. The first astronaut, dressed in a blue and yellow suit, is standing with his hands behind his back, looking at the camera. The second astronaut, also in a blue and yellow suit, is standing with his hands on his hips, looking at the camera. The third astronaut, wearing a beige suit with a patch on the shoulder, is standing with his hands on his hips, looking at the camera. The space station has a green wall in the background."
+    ],
+    "video_ids": [
+        "TY0wIV2y7Hs_3_51to204"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three astronauts in different colored suits.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three men in what appear to be astronaut uniforms, each wearing a different colored suit: one in beige, one in blue, and one in maroon. These suits are distinct in color, fulfilling the condition of 'Three astronauts in different colored suits.'\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three individuals dressed astronaut appear to be astronaut suits, each wearing a different colored suit: blue in blue, one in blue, and one in aoon. The suits are distinct in color and fulfilling the ' of havingdifferent astronauts in different colored suits.' The}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f818ae97f7f26e3dd41fed4fc440ba2fcbccd2d7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5c23f97dd842c82dd9e37dfa77cd964169a78100f8366814444b4edf73b5ce7
+size 188408
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..77ef1a40e3a061a7e42878cb58da0bd5f52874c9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is a football player, wearing a white and red uniform with the number 7, who is in the process of catching a football. The player is wearing a helmet and has tattoos on his arms. The football is brown and is being held securely in the player's hands. The player is in motion, running towards the camera, indicating that he has just caught the ball. The background is blurred, suggesting that the action is taking place on a football field with other players and spectators. The style of the video is action-oriented, capturing the intensity and excitement of the game."
+    ],
+    "video_ids": [
+        "G7mAXudIbos_4_180to334"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A football player in a white and red uniform with number 7, and a brown football.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a white and red uniform with the number 7, holding a brown football. The player's attire, including the jersey and helmet, matches the description, and the football is clearly visible in his hand. The presence of tattoos and a blurred background does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a white and red uniform with the number 7 on and a brown football. The player is attire and the the helmet and helmet, matches the description. and the football is clearly visible in the hands. The setting of a on the wrist background does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8b3504178b44ba8abff023ee2483cc5c7ca6050c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f719377bcec3bc88eb8a7abd5c80dd7179e59eb859c3dd63ecb92851cb7cfbb4
+size 274593
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..9425945ecb1ede55285a43d3f3f8b2912b7937be
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a small island in the middle of the ocean. The island is lush with green trees and surrounded by clear blue waters. In the first frame, a boat is seen approaching the island, its white hull contrasting with the deep blue of the ocean. The boat is moving towards the right side of the frame, creating a sense of anticipation.  In the second frame, the boat has reached the island and is now docked at a small pier. The pier extends out into the water, providing a safe landing spot for the boat. The boat is now on the left side of the frame, indicating that it has completed its journey.  In the third frame, the boat has departed from the island and is now moving away from it. The island is now on the right side of the frame, indicating that the boat has successfully completed its journey to and from the island. The clear blue waters and the lush green trees on the island remain constant throughout all three frames, providing a sense of continuity and harmony in the video."
+    ],
+    "video_ids": [
+        "wNcxp3yIsNY_9_0to139"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: - Boat? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a boat in the water, positioned near an island. The boat is visible throughout the frames, and its presence aligns with the 'Object(s)' condition of the video description. Additional elements like a hand pointing or the island do not contradict the presence of the boat.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a boat in the water, which in a island. The boat is visible in the frames, and its presence iss with the 'Object(s)' condition. the video..\"\n The elements such the clearrail towards a presence itself not contradict the presence of the boat.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e26e6160c4e9d55ab03de44eccf93cffc483d379
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f9012f15204ba74332b6745e09934db092d5c15e4d514f59de8f9f05e89e27f3
+size 101240
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..9f9f24fb9999d527a914122c23ed1161f8e88207
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a wooden toy gun with a wooden bullet holder and wooden bullets. The gun is placed on a wooden surface, and the bullet holder is next to it. The bullets are scattered around the gun and the holder. The style of the video is a stop-motion animation, where the gun is shown in three different positions: loaded, fired, and empty. The animation is smooth and the movements are realistic. The wooden texture of the gun and the holder adds a rustic charm to the video. The wooden bullets add a touch of whimsy to the scene. The overall effect is a playful and charming representation of a classic toy gun."
+    ],
+    "video_ids": [
+        "vYKGxJTXPMw_49_102to229"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Wooden toy gun, wooden bullet holder, wooden bullets? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a wooden toy gun with brass accents, a wooden bullet holder (powder flask) with engraved designs, and several wooden bullets (lead shot) scattered nearby. All core elements described in the condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a wooden toy gun, a-colored, a wooden bullet holder,whichder flask), with brass details, and several wooden bullets.car shot). scattered around. The these elements of in the question are present, match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a692b1c6ca75b98639ae3433e3e29c6ecd9f8c2b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1308-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06f09f8084244589c4e6451a18413c4a8bf36d874d29056f483ab49fdfc51dbf
+size 229059
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..5094c2b232e6e17bc3b52bb2facbac571648df00
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a drone shot of a rural landscape, showcasing a highway cutting through the middle of a vast, flat, and dry terrain. The highway is surrounded by sparse vegetation and a few buildings, indicating a small settlement or town. The sky is overcast, casting a soft light over the scene. The video captures the vastness and emptiness of the landscape, emphasizing the isolation of the highway and the small settlement. The style of the video is realistic, with a focus on the natural landscape and the man-made structures within it. The drone perspective provides a unique view of the landscape, allowing the viewer to appreciate the scale and isolation of the area."
+    ],
+    "video_ids": [
+        "55q2MjWKHWM_49_0to133"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Highway, sparse vegetation, buildings, small settlement? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a wide aerial view of a rural landscape with a visible highway running through it, sparse vegetation (dry grass and scattered trees), a few buildings (including what appears to be a farmhouse or small structures), and a small settlement in the distance. These elements align with the described conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a highway highway view of a highway landscape with a long highway running through it. which vegetation scatteredtrees fields and scattered trees), and few buildings,houses houses appears to be a farm and small settlement), and a small settlement. the distance. The elements collectively with the ' ',}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9320a744b45fa6aa765b4e3b747852d182ee3831
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b82cba31d8e46eaec8bb1184d04c8e9fc21ad35b86c67e60169eaa2d5f47338b
+size 160199
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..e16ebf46d6eddb8a84d102908fdbe451c5a5824c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up shot of a dish being prepared, focusing on the ingredients and the process of mixing them together. The dish appears to be a salad, with visible ingredients such as tomatoes, onions, and greens. The style of the video is simple and straightforward, with a focus on the food and the preparation process. The camera angle is close and direct, capturing the textures and colors of the ingredients as they are mixed together. The lighting is bright and even, highlighting the freshness and vibrancy of the ingredients. The video does not contain any text or additional elements, focusing solely on the food and the preparation process."
+    ],
+    "video_ids": [
+        "kCyER5ehAcw_40_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Tomatoes, onions, greens (ingredients of the salad).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a salad with visible ingredients including red tomatoes, white onions, and green leafy herbs, which aligns with the description of the salad's components. The visual details match the specified ingredients without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a close with visible ingredients that tomatoes tomatoes, white onions, and green leafy greens, which ares with the description of tomatoes salad's ingredients. The presence elements match the core ingredients, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..36e24c44f2f27ede0e426803fa44acce6759f257
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:faa41560ec6f551606a07830277927f14d2d150154a5bb414fe3410bdf7dcd25
+size 113551
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..884f79b7dcbdc650c4cda1dbb8195985d612363d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a bald man with a beard and mustache, wearing a red shirt and a backpack, standing in a forested area. He is pointing towards the camera with his right hand, and his expression is neutral. The forest is lush and green, with tall trees and dense foliage. The lighting suggests it is daytime. The man appears to be in the middle of a hike or outdoor adventure. The style of the video is candid and informal, capturing a moment in the man's journey."
+    ],
+    "video_ids": [
+        "2zUTn4uRkQQ_21_368to538"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bald man with a beard and mustache, wearing a red shirt and a backpack.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man with a beard and mustache, wearing a red shirt and a backpack, which matches the description. The background and other elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man with a beard and mustache, wearing a red shirt and a backpack. standing matches the description provided The background is the elements in not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..969d7e91a5f9e9af8317f065f9755c932c0173d1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c7739ddff3d0afd4b935c3e518180c236f594ef095ee0e3bbcc2ff154d07eb64
+size 252832
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..3fcc18a5b311adab4e1714bc78eb97a5e0599599
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a vintage car in a field, with a focus on the rear end. The car has a distinctive chrome bumper and tail fins, which are rusted and show signs of wear. The car is parked on a dirt field, with trees and bushes in the background. The video is shot in a realistic style, capturing the details of the car and its surroundings. The car appears to be in a state of disrepair, suggesting it has been abandoned or neglected for some time. The overall mood of the video is nostalgic and melancholic, evoking a sense of the passage of time and the decline of the car's once-glorious appearance."
+    ],
+    "video_ids": [
+        "ReZPLsInAhA_16_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Vintage car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a vintage car, specifically an old, rusted model with classic design elements such as tailfins and chrome details. The car's aged and weathered condition aligns with the 'vintage' description, and no elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features the vintage car, specifically focusing older model classiced model with a design elements such as a finsins and a detailing. The setting is rear appearance weathered appearance,s with the 'vintage' description. and the other contradict this core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e32bc7670f1031edba78449983b9ede27a5f8d50
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e74ba642547670d0efc65cf5cf2f9c59d8de91a8db4f57008accca46af2751c4
+size 158947
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..d4213c50795528d6d9ab32409209824032d0e220
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person making a small, colorful cupcake sandwich. The person is using a knife to spread frosting on the cupcake. The cupcake is placed on a black countertop. The frosting is yellow and green. The person is wearing a blue shirt. The cupcake sandwich is being made on a black countertop. The person is using a knife to spread the frosting on the cupcake. The frosting is yellow and green. The person is wearing a blue shirt. The cupcake sandwich is being made on a black countertop. The person is using a knife to spread the frosting on the cupcake. The frosting is yellow and green. The person is wearing a blue shirt. The cupcake sandwich is being made on a black countertop. The person is using a knife to spread the frosting on the cupcake. The frosting is yellow and green. The person is wearing a blue shirt. The cupcake sandwich is being made on a black countertop. The person is using a knife to spread the frosting on the cupcake. The frosting is yellow and green. The person is wearing a blue shirt. The cupcake sandwich is being made on a black countertop. The person is using a knife to spread the frosting on the cupcake. The frosting is yellow and green. The person is wearing a blue shirt. The cupcake sandwich is being made on a black countertop. The person is using a knife to spread the frosting"
+    ],
+    "video_ids": [
+        "FrbMaA_E0Ng_12_0to139"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ['person', 'cupcake', 'knife', 'frosting']? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person decorating cupcakes with frosting, which matches the 'Object(s)' condition. The person's hand is visible, cupcakes are clearly depicted, and frosting is being applied. There is no knife present, but since the condition allows for additional elements as long as they don't contradict the core description, this is acceptable.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person using a with frosting. which involves the 'person(s)' condition. The person is hands is visible, holding with present present, a a is being applied. The is a indication present in but the the condition only for the elements, long as they do't conflict the core description, the is acceptable.}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8630eee1f5a8dfbe8e4cfc6d5c5e1110a50c662b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1504c0d578acdb9ba4d321b5250ebe28cbc3190643af04b264b6b4a254325d6f
+size 195067
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..4b2b74b87a325acf48398a29311c0d92ffb4bc42
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic scene of a blue and white speedboat cutting through a calm lake. The boat, with its sleek design and vibrant colors, is the main focus of the video. It's moving at a high speed, leaving a trail of white water behind it, indicating its swift movement. The lake itself is vast and serene, with a clear blue sky above and lush green trees lining the shore in the background. The boat's reflection is visible on the water's surface, adding a sense of symmetry to the scene. The overall style of the video is dynamic and energetic, capturing the thrill of speedboat riding in a beautiful natural setting."
+    ],
+    "video_ids": [
+        "qVsWoV25h-A_0_138to262"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue and white speedboat? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a speedboat with a blue and white color scheme, matching the description. The boat is in motion on water, and its design and colors are consistent with the 'blue and white speedboat' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a blueboat with a predominantly and white color scheme, which the description. The boat is in motion on a, and the design and color are consistent with the 'blue and white speedboat' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7a60a1402cde85558330423fcf11963e087a6f53
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06f2cc17c048cdd72640430e9e6485dba9e4127186f3c42f524d16aede8b69aa
+size 214802
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..d33cb68fcabfbcbe78c1bd67151517462fa38f80
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a person's hand holding a yellow corn cob on a wooden table. The corn cob is fresh and ripe, with visible kernels. The person's hand is holding the corn cob with a wooden skewer inserted through it, suggesting that the corn cob is being prepared for grilling or roasting. The wooden table provides a rustic and warm background for the scene. The style of the video is simple and straightforward, focusing on the corn cob as the main subject. The lighting is natural and soft, highlighting the texture and color of the corn cob. The video does not contain any text or additional elements, keeping the focus solely on the corn cob and the person's hand."
+    ],
+    "video_ids": [
+        "4VyMFJ1qABo_64_0to123"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person's hand holding a yellow corn cob with a wooden skewer? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person's hand holding a yellow corn cob, and a wooden skewer is inserted into the corn. The core elements described in the condition are present and accurately depicted, with no contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person's hand holding a yellow corn cob with which the wooden skewer is inserted into the corn cob The description elements of in the question are present and accurately depicted in with no additional or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0ca0e315888f0f9bcadf3783cf3b6e2ee89d12b3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90663f38409cb6e5b05839d23254c1e2257d7c73b7dcf18b22a862f678046a3e
+size 125631
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..04b70893aa083b5c16a944f4c7e897dcfc93165d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a glass of orange juice on a white lace tablecloth. The glass is filled with a frothy orange liquid, and there is a slice of orange next to it. The orange slice is cut in half, revealing the juicy pulp inside. The glass is placed on a white tablecloth with a lace pattern. The background is blurred, but it appears to be a kitchen counter. The video is likely a tutorial or advertisement for a method of preparing orange juice."
+    ],
+    "video_ids": [
+        "ME5xo-mYVNM_11_0to191"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A glass of orange juice and a half-slice of orange.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two glasses of orange juice and a half-slice of orange on a tablecloth, which matches the core description. Additional elements like the text overlay and background do not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a main of orange juice and a half-slice of orange. a white.. which matches the description description provided There elements like the table ' and the do not contradict the description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ff92745e799b377fa59d0f23d64513da904fd9bd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1309-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:677ea11fc9a655c9311235497a4ffac6879b752e1a5d28b5ca2a5bb409cfe263
+size 102123
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..106fbd01eeda2232a8bb4a8921c164561d164337
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car from the perspective of the driver's seat. The car is a modern, luxury vehicle with a sleek design. The dashboard features a large touch screen display that shows a map, indicating that the car is equipped with GPS navigation. The steering wheel is on the left side of the car, which suggests that the car is designed for driving on the right side of the road. The car's interior is well-lit, with natural light coming in through the windows. The seats are upholstered in a light-colored material, and the car's interior design is minimalistic, with a focus on functionality and comfort. The car appears to be in motion, as suggested by the view of the road and other vehicles outside the car. The video is likely a promotional video for the car, showcasing its features and design."
+    ],
+    "video_ids": [
+        "IcYcwbypv_s_57_0to148"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dashboard with a large touch screen display, steering wheel on the left side, seats upholstered in light-colored material? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a dashboard with a large touch screen display, a steering wheel positioned on the left side, and seats upholstered in light-colored material. All core elements of the description are accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a dashboard with a large touch screen display, a steering wheel on on the left side, and seats upholstered in light-colored material. These these elements of the description are present represented in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..938b0032e9ae82128cfd7494203f6a11c6b7c87b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:323d13e285f124d59915e3f481ecb62bdc47404aa7e7f45946379f5b54527d9b
+size 199561
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..9de8e7e750010688ab7d9d79cfed5fc7e3f9d6e3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person preparing a drink using a blender. The person is pouring a liquid, likely coffee extract, into a glass filled with ice. The blender is placed on a countertop, and there are other ingredients nearby, including milk and chocolate powder. The style of the video is instructional, with text overlaying the image to provide information about the ingredients used in the drink. The text reads \"INGREDIENTS 3 oz. 100% Pure Coffee Extract\". The overall setting suggests a home kitchen environment."
+    ],
+    "video_ids": [
+        "BH1QOGDR5Ms_8_0to122"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person, blender, glass filled with ice, liquid (coffee extract), milk, chocolate powder? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person pouring coffee extract into a glass filled with ice, with a blender (Vitamix) in the background. A glass of milk and a container of chocolate powder are also visible, matching the described ingredients. All core elements are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person pouring a extract into a blender filled with ice, which additional blender andlikelyitamix) in the background. The glass of milk and a container of chocolate powder are also visible on which the objects objects. The elements elements are present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..aaee4ca9bfd6107b50d2002452e2bdb1a0eab239
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9d7c7b03583342999e36c868a377bbcbcbaa2268aea828a9e7415693b529cb86
+size 146303
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..3404dd6c9b61f4721c3187854d645841370534a6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and a brown jacket, sitting in a chair and gesturing with his hands as he speaks. The setting appears to be a television studio, with a blue background and a logo in the corner. The man seems to be engaged in a conversation or interview, and his body language suggests he is making a point or explaining something. The style of the video is a standard television interview or talk show, with a focus on the man's facial expressions and hand gestures to convey his message."
+    ],
+    "video_ids": [
+        "SsLEv8h4EtY_3_0to126"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and a brown jacket, sitting in a chair.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man with a beard wearing a brown jacket, seated in a chair. The description matches the visual content without contradiction, even though there are other elements in the background and another person partially visible.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man with a beard and a brown jacket, sitting in a chair. The background matches the visual content of any.\"\n and though the are no elements in the background, the person partially visible,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ed89144e5d7a0f40ff39dbcfb64c55c3583bf83b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:20ea7cf7bc492c6cdc6c6f6b80ca7618aa0520614836df05fe3812e1f588f62a
+size 129369
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..71c85dd606e0d3a662f2a0d8e998deee035c4e9f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures an underwater scene featuring a school of salmon swimming through a rocky riverbed. The water is clear enough to see the pebbles and stones at the bottom, which vary in size and color, ranging from light gray to dark brown. The salmon, with their distinctive reddish-orange bodies and silvery sides, swim in a coordinated manner, their fins and tails moving rhythmically as they navigate the current. Some fish are closer to the camera, while others are further away, creating a sense of depth. The water's surface is visible above them, with ripples and reflections indicating movement on the water's surface. The lighting is natural, suggesting the video was taken during daylight hours, and it highlights the vibrant colors of the salmon and the textures of the rocks below. The overall atmosphere is serene and natural, showcasing the beauty of these fish in their habitat."
+    ],
+    "video_ids": [
+        "936477b8d8c573905f09aadfac3366a74748ba4605a445c30fba9282379a712a"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A school of salmon with reddish-orange bodies and silvery sides.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple salmon swimming underwater, with their bodies displaying a reddish-orange hue and silvery sides, which matches the description. The underwater environment with rocks and clear water also supports the natural habitat of salmon. While some fish may appear more reddish or less distinct due to lighting or motion, the overall appearance aligns with the described characteristics.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a fish-like in, which their bodies exhibiting a reddish-orange color and silvery sides, which align the description provided The presence environment with pe and the water further supports the setting habitat of salmon. The there additional appear appear slightly orangeish than sil sil, to the or angle, the overall depiction aligns with the given ' of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..43b9810609b2b741af4e085008294308ebbb2700
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d1dd88930afdb81bc42445376f1c4759effa2c3c61711e54a6384aa4fc7ea215
+size 240771
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..3c5e197008fef59d90ca132919e4afd5a6f1061f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman with long, dark hair and glasses is seen in a dark room. She is wearing a camouflage jacket and is seated in a chair. Her head is tilted back and she is looking up at the ceiling, her mouth open as if she is in the middle of a conversation or reacting to something. The room is dimly lit, with the woman being the main focus of the image. The style of the video is candid and informal, capturing a moment in the woman's life."
+    ],
+    "video_ids": [
+        "Ouae4ILzy48_21_0to154"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with long, dark hair and glasses, wearing a camouflage jacket, sitting in a chair.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with long, dark hair and glasses, wearing a camouflage-patterned jacket, seated in what appears to be a chair. The core description is accurately represented, with no significant contradictions. Additional elements (e.g., background people, slight camera movement) do not conflict with the primary subject description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a person with long, dark hair and glasses, wearing a camouflage jacketed jacket, sitting in a appears to be a chair. The setting elements is largely represented in with no significant contradictions or The elements,like.g., the details, lighting changes movement) do not detr with the main description and.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ebecfe251c5eabbd59b0f447f7d77c04812b4a43
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dc40d614f8a44fe4fa1909c98f51b009939ded4e9d3acb564c2746a777b6156b
+size 108971
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..96d41d889cb1a38571442708624b32746e1ef90b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young boy with blonde hair, wearing a blue jacket with white stripes, sitting at a table in a classroom setting. The boy appears to be engaged in a conversation or an interview, as he is looking directly at the camera with a serious expression on his face. The classroom has a window with blinds, and there is a desk with papers on it in the background. The lighting in the room is bright, and the overall atmosphere is calm and focused. The style of the video is straightforward and documentary-like, with no additional effects or embellishments."
+    ],
+    "video_ids": [
+        "7YU0ix7xloQ_32_0to159"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Young boy with blonde hair, wearing a blue jacket with white stripes? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young boy with blonde hair wearing a blue jacket with white stripes on the sleeves and shoulders, which matches the description. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young boy with blonde hair, a blue jacket with white stripes. the sleeves. a. which matches the description provided The setting and additional elements in not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e750fcc600ce8b935dc314b84536be61f2dc277e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:efa84f0f2175e8a86fee6b9fa55b83f065cbd9d1782b4ced33396854551fe805
+size 168905
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..52346d80b9ced346b7ee8fe882ae37dda7be6521
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a collection of brown shoes, each with a unique design and style. The first shoe is a classic brown oxford, featuring a smooth finish and a laced front. The second shoe is a brown loafer, characterized by its slip-on design and a strap across the top. The third shoe is a brown monk strap, distinguished by its single strap and buckle closure. The shoes are displayed on a marble surface, which contrasts with the brown tones of the shoes. The video emphasizes the importance of ensuring that brown tone shoes are different from other shoes, highlighting their distinct features and styles."
+    ],
+    "video_ids": [
+        "SQazl-HSYAE_53_0to188"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Brown shoes (Oxford, Loafer, Monk Strap)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video displays three distinct pairs of brown shoes: an Oxford (laced), a Loafer (slip-on with a strap), and a Monk Strap (with two buckles). These match the specified object types and color, fulfilling the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two brown types of brown shoes, an Oxford shoetopaced shoe a Loafer (withip-on with a buckle), and a Monk Strap (sl a strapsles). These types the description ' types, the, fulfilling the '.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4893835322845f037cde97a05da63aa742c03911
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4a73e38d8165b1675bce9ef07f5f1609c7fc639370d8573079b2808d81e8cb52
+size 89765
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..c993a280db9b76ed7852b6413b27797530effa1d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen preparing food in a kitchen. He is wearing a black shirt and glasses. The kitchen is well-equipped with various appliances and utensils. On the counter, there are several potted plants adding a touch of greenery to the space. The man is using a waffle maker to cook waffles, and he is also frying chicken on a pan. The video captures the process of food preparation, showcasing the man's culinary skills. The overall style of the video is casual and homey, with a focus on the man's actions and the food being prepared."
+    ],
+    "video_ids": [
+        "coHZJx_1-Jw_51_118to278"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a waffle maker, a pan, several potted plants? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man using a waffle maker to remove a cooked waffle with tongs. A pan (specifically, a metal baking sheet with food on it) is visible in the background, along with several potted plants on the countertop. All core objects mentioned in the condition are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man in a waffle maker and cook a w waffle. aongs. There pan islikelyally a a frying frying pan) a remnants it) is also on the background. and with several potted plants on the kitchenop. The the elements are in the condition are present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5d74c88c50308b84717dba6c02d169a711d3822d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-131-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0655b7f8d37779c830dff26ab117e6d80a0db7a565bf72632236eef33804a613
+size 237401
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..06229634c367df73f6a24b9c299368ce526de994
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a black Nintendo Wii U game console and its controller, the Wii U GamePad, being used on a wooden table. The console is placed on the table, and the GamePad is placed on top of it. The GamePad is connected to the console via a cable. The console is turned on, and the GamePad is displaying a blue screen with the Nintendo logo. The video captures the setup of the console and the controller, and it shows how they are connected. The video is likely intended to demonstrate the setup process for the Wii U console and its controller."
+    ],
+    "video_ids": [
+        "fgOxHlYCeuc_1_119to272"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Nintendo Wii U game console, Wii U GamePad? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a Nintendo Wii U game console with its GamePad resting on top of it. Both objects are identifiable and match the description provided, with no conflicting elements present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows shows a Nintendo Wii U game console with a GamePad attached on top. it. The the are visible by the the description.. with the additional elements present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..736e59299c0d82774a3931978c7ce93520234749
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26c888867a670ce053fb0885ecbe96dec0d23558beeb132d8f3b9e8cd1cc6155
+size 208203
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..d23ba5e5706e97407c19aceab639fe18597d12c1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in front of a blue curtain. He is wearing a blue shirt and has a beard. His hands are raised, and he appears to be speaking or gesturing. The style of the video is casual and informal, with the man being the main focus. The blue curtain provides a simple and uncluttered background, allowing the viewer to focus on the man and his actions. The lighting in the video is bright, suggesting that it was filmed during the day. The overall impression is that of a friendly and approachable individual, possibly engaging in a conversation or presentation."
+    ],
+    "video_ids": [
+        "1hSrC227IF8_0_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a blue curtain? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man standing in front of a blue curtain, which matches the 'Object(s)' condition. The man is the central subject, and the blue curtain serves as the background, fulfilling the description without any conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man wearing in front of a blue curtain. which ful the 'Object(s)' condition. The man is wearing primary figure, and the blue curtain serves as the background, fulfilling the description provided any contradictions elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4c1336ac488156904feddd53daf5474278b1a79c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1cd29d88a6def53d09177650c40fb9e16278738072a607768cab72e73eaed9f6
+size 172440
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..f49ed957d41bb332874d4366d7efc01308e1221c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a 2019 Chevrolet Blazer in a spacious, well-lit warehouse. The car is positioned in the center of the frame, facing the camera, and is captured in three different angles. The first angle is a front view, highlighting the car's sleek design and prominent grille. The second angle is a side view, emphasizing the car's aerodynamic shape and alloy wheels. The third angle is a rear view, showcasing the car's taillights and rear bumper. The warehouse setting provides a stark contrast to the car's vibrant red color, making it stand out prominently. The video is a professional car advertisement, designed to showcase the car's design and features."
+    ],
+    "video_ids": [
+        "UEFbsoSTCQ4_3_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: 2019 Chevrolet Blazer? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red 2019 Chevrolet Blazer, as indicated by the text overlay in the bottom left corner. The vehicle's design, including the front grille, headlights, and overall shape, matches the known appearance of the 2019 Chevrolet Blazer model. There are no visual elements that contradict this identification.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a red Chevrolet2019 Chevrolet Blazer in as indicated by the distinct and and the image image corner of The vehicle is design, including its front grille, headlights, and overall shape, matches the description characteristics of the 2019 Chevrolet Blazer.. The are no additional elements that contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c36ec24e7d3440b2362a9058b7a478b132ead4cd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2ddbe1acfd9fef69729ce5bad1810d28993eaad5814ca0f9f5f9c483ec2d5b03
+size 151244
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..71f9169c4673ef24173e7092b9a36d0ac163e026
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a coffee shop display featuring Starbucks coffee. The display consists of two wicker baskets filled with coffee bags. The baskets are placed on a table, and there are signs in front of each basket. The signs are in Spanish, indicating the type of coffee available. The coffee bags are neatly arranged in the baskets, and the display is well-organized. The coffee shop appears to be located in Indonesia, as indicated by the sign. The style of the video is straightforward and informative, focusing on the coffee display and the available options for customers."
+    ],
+    "video_ids": [
+        "nrC6jh57kII_25_21to211"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two wicker baskets filled with coffee bags, signs in front of each basket, coffee bags in the baskets.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two wicker baskets filled with coffee bags, each with a sign in front. The signs describe the coffee types, and the baskets contain various coffee bags as described. There are no elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two wicker baskets filled with coffee bags. with with a sign in front of The signs appear the contents bags, and the baskets are coffee coffee bags with expected. The are no additional in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..de0e877472ba262e8706ea3ee8ac474f1925e6bc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:999178c9c41811ec0174a7607c47d3732a2281890eda548cddd1760fef7e55c1
+size 223658
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..3bcd5db344af73a605b75fff95e2290b5160e51c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up of a plant with white flowers and green leaves. The plant is in focus and the background is blurred. The flowers are in bloom and the leaves are lush. The plant is growing in a garden and the garden is well-maintained. The plant is the main subject of the video and there are no other objects in the video. The style of the video is a close-up nature shot."
+    ],
+    "video_ids": [
+        "h_2GMByKxNQ_60_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A plant with white flowers and green leaves, which is in focus and blooming.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a plant with white flowers and green leaves, which is clearly in focus and appears to be blooming. The flowers are detailed and vibrant, matching the description. While there is a bee and text overlay, these elements do not contradict the core description of the plant.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a plant with white flowers and green leaves, which are in in focus and appears to be blooming. The flowers are the and the, and the description of The there may a slight visible some visible in these elements do not contradict the core description of the plant and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7a564f499be567c07a2e43d088913fad76c666c1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:265eee8142ca7bcee569ef0cc316ea88647f23cb9bc993b950b75ff18dfcf873
+size 184970
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..a21a1f52c8226fe1e033a7c86adf4e7c668f5867
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a cooking tutorial that features a close-up view of a meal being prepared. The style of the video is simple and straightforward, focusing on the ingredients and the process of cooking. The video begins with a shot of a bag of chicken strips, followed by a bottle of chili lime seasoning, and then a bottle of onion salt. The ingredients are arranged on a pink cutting board, which is placed on a table. The camera captures the details of the ingredients, including the texture of the chicken strips and the color of the seasoning. The video does not include any text or narration, relying solely on the visuals to convey the cooking process. The overall style of the video is minimalistic, with a focus on the ingredients and the cooking process."
+    ],
+    "video_ids": [
+        "aaqZzNydL10_1_270to428"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bag of chicken strips, bottle of chili lime seasoning, bottle of onion salt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a bag of chicken strips, a bottle of chili lime seasoning, and a bottle of onion salt, all of which are explicitly mentioned in the 'Object(s)' condition. These items are prominently displayed and identifiable in the frame, fulfilling the requirement.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows shows a bag of chicken strips, a bottle of chili lime seasoning, and a bottle of onion salt. which of which are mentioned mentioned in the questionObject(s)' condition. The items are present displayed in are in the video, fulfilling the condition of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f53fe832d53323e7d9b3e3f651e0c7657ff751e1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75fd62f56021ace9d3fdc4c98f5522367c1edb9bd32ea64ed69c3bd37266ed6b
+size 199540
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..22dde1531307a7b88301ea01f5a51198ebc501e4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a police officer is seen in a city street, leaning against a tree. He is wearing a blue helmet and sunglasses, and is holding a walkie-talkie to his ear. The officer is dressed in a black jacket and khaki pants. In the background, there are cars parked on the street, and a stop sign is visible. The scene suggests that the officer is on duty, possibly monitoring traffic or responding to a call. The overall style of the video is realistic, capturing a typical moment in the life of a police officer in an urban setting."
+    ],
+    "video_ids": [
+        "B_bseKkpRi4_17_0to172"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A police officer, a tree, a walkie-talkie, a blue helmet, sunglasses, a black jacket, khaki pants.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person wearing a blue helmet, sunglasses, a black jacket, and khaki pants, leaning against a tree. While a walkie-talkie is not clearly visible, the person's attire and posture suggest they might be a police officer. The background includes a tree and cars, consistent with the description. The core elements are largely present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a person wearing a blue helmet, sunglasses, a black jacket, and khaki pants, which against a tree. The the walkie-talkie is not visible visible, the person appears hand and posture suggest they are be using police officer. The presence includes a street and a, which with the '. The presence elements of present present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f9ad1894c6e3e0cb02c884a6779d769ffc1f0075
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0ba80a4790d44e87d086f6c057cf7e06beedbf733e366598eb93acf888cb0feb
+size 165304
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..17bfa6799b4b015c7b3a6b86cbe11559e49ee2d3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a soccer player in action, wearing a blue jersey with white stripes and a red and white logo on the front. He is seen in three different positions, running across the field with intensity and focus. The background is a blur of spectators, indicating a live game scenario. The player's movements and the crowd's presence create a dynamic and energetic atmosphere."
+    ],
+    "video_ids": [
+        "siwD6kMjnIk_28_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A soccer player wearing a blue jersey with white stripes and a red and white logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a blue jersey with white stripes on the shoulders and a red and white logo on the chest, which matches the description. The jersey also features the sponsor 'VEOLIA' and other minor details like armbands and tattoos, which do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a blue jersey with white stripes and the sleeves and a red and white logo on the chest. which matches the description provided The player is has a word 'SGAIA' in the details details that thembands and socks, which are not contradict the core description.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f31886c440e2aafe4f82c83f46e3db7f26c68949
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1310-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:35f10b9df5b254599cbbbb0ec3ebb1d592c04d8645ef301b9dfe3f5687603689
+size 339529
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..1d846ca73793272242b61ee4a8bd406a644cbd2f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a jubilant moment of a football team celebrating their victory. The team, dressed in their uniforms, is seen in a group hug, with one player holding the championship trophy high above their heads. The atmosphere is filled with joy and excitement, as confetti falls around them, adding to the festive mood. The team's expressions are of pure happiness and pride, reflecting the hard work and dedication that led to this moment. The video is a testament to the spirit of teamwork and the thrill of victory in sports."
+    ],
+    "video_ids": [
+        "kS-TjF6lNRA_41_149to270"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A football team in their uniforms, a championship trophy? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group of football players in their uniforms celebrating with a championship trophy. The players are wearing team jerseys, and one player is holding up the trophy. Confetti is falling, which is consistent with a post-game celebration. The core elements described \u2014 a football team in uniforms and a championship trophy \u2014 are clearly present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group of individuals players in their uniforms celebrating with a championship trophy. The players are wearing light jerseys, and the player is holding the a trophy, Theetti is falling around which is a with a celebr-game celebration. The scene elements of in a football team in their and a championship trophy \u2014 are present present in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..238679017a237621956342e835f73a170c357b5b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6499d385e5ee212629fe673fecfaefaa88fc40a08a92b14d2c982c98c2690d56
+size 292911
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..6947a491ced8f3aef8ea311e730b5ffa94775dfe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is an aerial view of a park with a variety of landscapes. The park features a large open field, a dirt area, and a river running through it. The field is lush and green, and it's surrounded by a network of paths and roads. The dirt area is in the center of the park, and it's surrounded by trees. The river is calm and meandering, and it's lined with trees and shrubs. The park is located in a suburban area, and there are houses and roads visible in the background. The video is shot in daylight, and the colors are vibrant and clear. The style of the video is realistic and naturalistic, with a focus on the beauty of the park and the surrounding environment."
+    ],
+    "video_ids": [
+        "yrH3TywyrcI_18_0to145"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ['large open field', 'dirt area', 'river', 'trees', 'houses', 'roads']? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a large open field (green grassy areas), a dirt area (brown patch of land), a river (curving water body), trees (dense green foliage), houses (visible in the distance), and roads (paved pathways and streets). All specified elements are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a large open field,green areay area), a dirt area (a patch in land), a river (bodyved body body), trees (green greenery), houses (build in the background), and roads (curaved paths). streets). The these elements are present and contribute depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d07ffdc45c04838455101c57d93e8d31694e36fb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37af61e879d21c564e8cb6d5fd588a5fb588a13b2d1bdea0cc7cf74526596cce
+size 191945
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..ab6865d278911521c05efebb58a2ffecc68943d6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the process of making a dessert, specifically a cream puff. The first frame shows a cream puff being filled with whipped cream from a piping bag. The second frame shows the cream puff being placed on a cooling rack. The third frame shows the cream puff cooling on the rack. The style of the video is a close-up, time-lapse shot that focuses on the details of the dessert-making process. The video does not contain any text or additional elements. The focus is solely on the dessert and the process of making it."
+    ],
+    "video_ids": [
+        "m-v6vAHkYrA_33_204to392"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Cream puff, piping bag, cooling rack? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows cream puffs on a cooling rack, with a piping bag being used to fill one of them with cream. All specified objects are present and correctly depicted without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a beinguffs being a cooling rack, with a piping bag being used to fill them of the with cream. The the objects are present and the depicted in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1c617004650cc268947e57335c0b0f97da158d19
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87c02ee6677a2fb638e9dd8181d258cf7cbda06044db408ed89beccee6993fa4
+size 91774
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..e8b78f562437ef22bc41cac8c30dfa8857bc25af
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen in a garage, holding a black cloth over his eyes. He is wearing a gray t-shirt and appears to be in the process of cleaning or polishing a car. The garage is filled with various tools and equipment, suggesting that it is a space for working on vehicles. In the background, there is a red car and a motorcycle, indicating that the man might be a car enthusiast or a mechanic. The style of the video is casual and informal, capturing a moment in the man's daily routine."
+    ],
+    "video_ids": [
+        "_9yTXjXc3Ko_19_323to491"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a black cloth, a gray t-shirt, a car, a motorcycle? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a gray t-shirt, holding a black cloth, and standing in front of cars (including a black car with its hood open and a red car in the background). There is no motorcycle visible in the video, but since the description allows for additional elements as long as they don't contradict the core description, the absence of a motorcycle does not invalidate the fulfillment of the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a gray t-shirt, holding a black cloth in and standing in front of a andincluding a red motorcycle and a hood open and a red car). the background). There is also motorcycle visible in the video, but the the question does for additional elements, long as they do't conflict the core description, the presence of a motorcycle does not dis the fulfillment of the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..745163c9b4af8a015b5837ee2359eca2b72ca6b2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:612abbdfd0189c339048004b00e39c22bb7c3ad817c3f0d894e1ab794b520c67
+size 168813
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..cfd22b6aaeeba77836c0eb9a03705ea1912af43d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures an underwater scene featuring a school of salmon swimming through a rocky riverbed. The water is clear enough to see the pebbles and stones at the bottom, which vary in size and color, ranging from light gray to dark brown. The salmon, with their distinctive reddish-orange bodies and silvery sides, swim in a coordinated manner, their fins and tails moving rhythmically as they navigate the current. Some fish are closer to the camera, while others are further away, creating a sense of depth. The water's surface is visible above them, with ripples and reflections indicating movement on the water's surface. The lighting is natural, suggesting the video was taken during daylight hours, and it highlights the vibrant colors of the salmon and the textures of the rocks below. The overall atmosphere is serene and natural, showcasing the beauty of these fish in their habitat."
+    ],
+    "video_ids": [
+        "936477b8d8c573905f09aadfac3366a74748ba4605a445c30fba9282379a712a"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A school of salmon with reddish-orange bodies and silvery sides.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple salmon swimming underwater, with their bodies displaying a reddish-orange hue and silvery sides, which matches the description. The underwater environment with rocks and clear water also supports the natural habitat of salmon. While some fish may appear more reddish or less distinct due to lighting or motion, the overall appearance aligns with the described characteristics.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a fish with in, with their bodies displaying a reddish-orange color and silvery sides, which align the description provided The presence environment, colorful and water water also supports the setting habitat of salmon. The there additional appear appear slightly vividish than sil sil, to the or angle, the overall depiction aligns with the given ' of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f62a26d9d6aaaaab941532a9e7590963bb87b1e4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6af17dc94d3b4aafef82e32acead448b0f7768d153aa038285c7f0c4f8ee0c58
+size 364445
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..cc9368a88b7d46d00c02c81c6dd960656055ae27
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a conversation between two men in a car showroom. The man on the left, dressed in a black suit, is holding a car engine, while the man on the right, wearing a blue suit, is gesturing with his hands. They are standing in front of a red car on display. The showroom is filled with other cars and people, creating a lively atmosphere. The focus of the video is on the interaction between the two men and the car engine, suggesting a discussion about the car's performance or features. The video style is candid and informal, capturing a real-life moment in a car showroom."
+    ],
+    "video_ids": [
+        "WdqTA3EOV2M_19_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men in suits, one holding a car engine and the other gesturing with his hands.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men in suits interacting with a car engine. One man is holding the engine while the other is gesturing with his hands, which matches the described condition. Additional background elements like people and cars do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men in suits, in a car engine. One man is holding the car, the other is gesturing with his hands, which align the description scenario. The elements elements like the and cars in not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f8dd19d6b0d73d0e54faf8d5f57a78d356fc8809
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb2960346bdd031e0a02fbfa5b8e98d36d32fc4795bb1c57857caf04076e5b96
+size 176555
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..22a2dd2df378fb3158c66d467e7e520852d7c99b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a kitchen, eating food from a bowl. He is wearing glasses and a gray robe. The kitchen is cluttered with various objects, including a green bottle, a red kettle, and a white bowl. The man is using a spoon to eat the food. The video is likely a casual, everyday scene, capturing a moment of the man's life. The style of the video is realistic and unfiltered, with no special effects or artistic alterations."
+    ],
+    "video_ids": [
+        "7XP7SAGkCPA_37_38to253"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a bowl of food, a spoon? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting at a table eating, with a bowl of food in front of him. He is using chopsticks (not a spoon) to eat. While a spoon is not present, the core elements \u2014 a man and a bowl of food \u2014 are clearly visible, and the presence of chopsticks does not contradict the 'Object(s)' condition as it is still a utensil used for eating. The description is largely fulfilled.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in in a table, from with a bowl of food in his of him and He is using asticks towhich a spoon) to eat from The the spoon is not present, the core elements of a man, a bowl of food \u2014 are clearly visible and fulfilling the action of chopsticks does not contradict the descriptionObject(s)' condition. it is a a formil for for eating. The red does largely matched.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dcca04e20da3baf217f14a3883a43fa75e7eb0a6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:140b3fc3c0939947fbd236a51677ea1e8f77b1b0fd98a821441c75b4dedc1563
+size 113696
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..1e3727ab03987ef4c6c2385c1ce9ce7fb780c571
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man sitting at a table in a restaurant. He is wearing a black tank top and has a beard. In the first frame, he is pointing at the camera with his right hand. In the second frame, he is holding a straw in his left hand. In the third frame, he is holding a jar in his right hand. The table has a blue tablecloth and there are several bowls on it. The restaurant has a wooden floor and there are other tables and chairs visible in the background. The man appears to be enjoying his time at the restaurant."
+    ],
+    "video_ids": [
+        "azjnyoMDFWA_0_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a straw, a jar? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man sitting at a table with a jar containing a straw. These core objects are present and accurately depicted, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man sitting at a table with a straw in a straw. The elements elements are present and match depicted. fulfilling the 'Object(s)' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0cbc4944edcbb52cf2b5795462c996d375ed7e6e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1311-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88845847596d4b07896a5cc125b56d2275ca15897455f96fcc34b8f9bfc31dfc
+size 118767
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..3b3d98398655a148685e01db2da1157ee0a2e95d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a baby crying in a close-up shot. The baby's face is the main focus of the video, with the camera capturing the baby's facial expressions and emotions. The baby's mouth is open wide, and the eyes are closed, indicating a strong emotional response. The baby's skin is pale, and the baby is wearing a yellow and white striped shirt. The background is blurred, but it appears to be a white surface, possibly a bed or a changing table. The style of the video is realistic and naturalistic, capturing a candid moment in the baby's life. The video does not contain any text or additional elements."
+    ],
+    "video_ids": [
+        "EnIFTPnu8Ys_3_0to178"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A baby? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a baby's face, which is clearly identifiable as a baby based on its facial features, skin texture, and overall appearance. The baby is crying, which is a common behavior for infants. The presence of the 'TOMONEWS' watermark does not contradict the object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a baby's face, which align the visible as a baby. on the features features, including texture, and the appearance. The baby is wearing, which is a common behavior in infants. The presence of a babybabyummyS'' watermark does not contradict the description condition as}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0bca8be6c809f352c476fbd8e07eb6c55041aa8f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6a13e973968b3256b20fa3b1c6adf5c7de53aa40311177921057cdafa1ebc04f
+size 93423
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..339e40fcf23cdea57712a00b08c79bd7e04d196f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man on a boat, holding up a smartphone displaying a speedometer reading of 21 mph. The man is wearing a cap and glasses, and he appears to be in motion, as suggested by the speedometer reading. The background features a body of water, possibly a lake or river, with trees and a cloudy sky. The style of the video is casual and seems to be taken from a first-person perspective, capturing the man's experience of speeding on the water."
+    ],
+    "video_ids": [
+        "rnp9JZSMC8Q_22_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man on a boat, a smartphone displaying a speedometer reading of 21 mph.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man on a boat, holding up a smartphone that clearly displays a speedometer reading of 21 mph. The man is wearing appropriate attire for being on a boat, and the background shows water and trees, consistent with being on a boat. The core elements described are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man on a boat, as a a smartphone that displays displays a speedometer reading of 21 mph. The background is wearing sunglasses attire for bo on a boat, and the background suggests a and a, which with a on a boat. The smartphone elements of in present represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ddf98e3005bc92038455d5b9f3d5775bcd2bbf5b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:404ba7223e7f268d81437210c236b767e518cfcead2ec2caa11bf0b6d3b8daf7
+size 153804
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..4c92f455707f9ac956af36c0de5fab5eb6b2a5a3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a car, focusing on the steering wheel and dashboard. The steering wheel is black and features the Jaguar logo in the center. The dashboard is equipped with a digital display and two analog gauges. The car's interior is well-lit, with the sunlight streaming in through the windshield. The steering wheel is positioned in the center of the frame, with the dashboard visible in the background. The video is shot from the perspective of the driver, providing a clear view of the car's controls and instruments. The style of the video is straightforward and informative, providing a detailed look at the car's interior."
+    ],
+    "video_ids": [
+        "p9K7Fh1tABM_9_0to117"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, dashboard (digital display and two analog gauges), Jaguar logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the steering wheel with the Jaguar logo prominently displayed in the center. Behind the steering wheel, the dashboard is visible, featuring two analog gauges (speedometer and tachometer) and a digital display area. All elements mentioned in the description are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows shows a steering wheel with the Chrysler logo, displayed in the center. The the steering wheel, there dashboard is visible, featuring a analog gauges andspeedometer and tachometer) and a digital display.. The elements match in the condition are present in match depicted in}<|im_end|>\n addCriterion",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c2a1637eb23f6f12f6265870b4085a6389052023
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:697bdfa672b6df892f50ec3979952234e71cc40a5bdad0fe08dafdf1a9b72f19
+size 197491
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..c8fe72ce9bde1b6a2d09c5bc8b0fd8301c08255d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dramatic scene set outdoors, featuring a large white cylindrical object on the left side of the frame. In the center, there is a wooden structure with a broken window, through which a dark silhouette of an animal, possibly a dog or a cat, can be seen. The ground around the structure is littered with debris and rocks. To the right, a small firework is ignited, emitting bright sparks and smoke. As the firework explodes, it creates a powerful burst of light and smoke, illuminating the surrounding area. The explosion is intense, with flames and smoke billowing outwards, engulfing the wooden structure and the debris on the ground. The firework continues to emit sparks and smoke even after the main explosion, creating a dynamic and visually striking sequence. The text \"210,000FPS\" appears at the bottom left corner, indicating the high-speed filming technique used to capture the event."
+    ],
+    "video_ids": [
+        "0a4c6460189eca8429bc2e0a4788837ae560a7c7b4e358ff2a9348653d956f71"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Large white cylindrical object, wooden structure with a broken window, debris and rocks on the ground, a small firework.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large white cylindrical object (likely a barrel) on the left, a wooden structure with a broken window in the center, debris and rocks on the ground, and a small firework-like explosion occurring near the right side. These elements match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a large white cylindrical object,likely a pole), next the left side a wooden structure with a broken window in the center, debris and rocks scattered the ground, and a small firework that object in inside the window side of The elements match the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..294ac614bda420922fa180f868950afaeacd24ba
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab48c70d82f9866e4d9d16fb35e15aa3e90d369285e846ecd60d611af0543465
+size 162459
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..b476c6cf5ae2c7c4f18a20a0068de10744128f35
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with blonde hair, wearing a dark vest over a red shirt. He is standing in front of a brick wall with a clock mounted on it. The man appears to be speaking, as suggested by his open mouth and attentive expression. The style of the video is a straightforward interview or conversation, with the man as the central figure. The setting is simple and uncluttered, focusing the viewer's attention on the man and his words. The lighting is even, with no harsh shadows, suggesting an indoor environment. The clock on the wall adds a touch of character to the scene."
+    ],
+    "video_ids": [
+        "WiFuFTjlhqI_11_0to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with blonde hair wearing a dark vest over a red shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The man in the video has blonde hair and is wearing a dark vest over a red shirt, which matches the description. The background elements, such as the brick wall and bicycle wheel, do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video in the video has blonde hair and is wearing a dark vest over a red shirt, which matches the description provided The background and, such as the brick wall and clock,, do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..561fda77d4937d446cbf52ba3c357cbed6695542
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:874375a7676e8edab729cad895aee0d0d39a77ded919143421deb75c572e3493
+size 106698
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..7de500068b05557d243cfe1f51756744e800237d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a small bird perched on a tree branch. The bird, with its blue, orange, and white plumage, is the main focus of the video. It is seen in three different positions on the branch, each time facing a different direction. The background of the video is filled with green leaves and branches, providing a natural and serene setting for the bird. The bird's movements and the changing angles of the shots create a sense of dynamism and life in the video. The overall style of the video is a close-up nature shot, focusing on the bird and its immediate surroundings."
+    ],
+    "video_ids": [
+        "8JtlUb-xujw_66_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small bird with blue, orange, and white plumage.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small bird with distinct blue (on the head), orange (on the chest), and white (on the wings) plumage, matching the description. The bird is perched on branches with some green leaves and pine cones in the background, which does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two small bird with blue blue,on the head and orange (on the chest), and white (on the belly and plumage, which the description provided The bird is perched on a with green greenery in the needles, the background, which is not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..370db0fa4bd55da0fe489d1d0659d0ebe6aaab1c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60ab54d19ee3fee3dfe4f709cbcda68a39fb0ced7b0111790642b9066f972e67
+size 226658
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..5e51abd8b2d8ae8a3b7c22898c865a17d1f41a43
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen standing in front of a building with barbed wire on top. He is wearing a black hoodie with a gold emblem on the front. The man is pointing his finger towards the camera, and his expression is serious. The building behind him has a window and an air conditioning unit. The overall style of the video is realistic, with a focus on the man and his interaction with the camera."
+    ],
+    "video_ids": [
+        "FZH42y1mVDg_233_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a building? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing glasses and a hoodie, pointing and speaking, and a building with windows and barbed wire in the background. These elements match the 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing a and a black, standing towards standing, which there building in windows and abed wire in the background. These elements fulfill the 'Object(s)' condition of any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b9398d443ee6377922fbe853cc0d6ea76c7d1e61
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9963a1e653af14c5d50033ae943e10cf26bdab72f95719fcb6720dbdf8186e53
+size 137418
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..50c39edf0fc6abc840741cbb369791efd6e434ff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man named Kevin Feige, a producer, who is being interviewed at a convention. He is wearing a black baseball cap and a dark suit. The background is filled with a crowd of people, suggesting that the event is well-attended. The style of the video is a standard interview format, with the interviewer's questions not visible in the image. The focus is on Kevin Feige, who is the main subject of the video. The overall atmosphere of the video is professional and serious, as it is likely discussing the production of a movie or other media."
+    ],
+    "video_ids": [
+        "UrNdFnN3y2Y_28_0to164"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Kevin Feige (a man wearing a black baseball cap and a dark suit)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black baseball cap with red embroidery and a dark suit, which matches the description of Kevin Feige. The lower third graphic also identifies him as 'KEVIN FEIGE, P.G.A. PRODUCER', confirming his identity and role.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black baseball cap and a and and a dark suit, which align the description of Kevin Feige. The background part of in suggests the as KevinKevinVIN FEIGE', CHRES.M.', PRESDUCER', confirming the identity. role.}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..339d83129446ebe7e44a337e0d189d13e20444d8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1312-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9617ed9757f914d6334f1918b9b7609f343f9b08bc784d7ff359b1bfc39c95c4
+size 142342
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..7432e0374c87ad02f2af6ba10cb3ffe7f7f78f4d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the lively scene of a carousel in motion, featuring a variety of colorful horse and pony figures. The carousel is adorned with golden poles and string lights, creating a festive atmosphere. As the carousel spins, the horses and ponies move up and down, mimicking the galloping motion of real animals. The riders, dressed in casual attire, hold onto the handles attached to the horses, enjoying the ride. The background shows a park-like setting with trees and other visitors walking around, adding to the cheerful ambiance. The camera remains stationary throughout the video, focusing on the carousel's movement and the joyful expressions of the riders."
+    ],
+    "video_ids": [
+        "d98dea6460af56db15c39d190782633632667b9a2e775b97f29fc461662f3be8"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Carousel with colorful horse and pony figures, riders holding onto handles, trees, and other visitors.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by showing a carousel with colorful horse and pony figures, riders holding onto handles, and trees and other visitors visible in the background. The carousel is rotating, and the riders are clearly interacting with it, matching the described elements without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting a carousel with colorful horse and pony figures. riders holding onto handles, and trees in other visitors in in the background. The presence is the, and the riders are depicted depicted with the, which the description elements.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dbdf6d5b2f702ceb58cc168bbd8284d2748ac805
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d640e572cbf6bb841ff62637223c7dff4d6408490bf1fe8fb09af1ae72480a2
+size 218449
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..a0c84e3f5cf496fb9dd9d710a38e653c0929b9e9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a hand with a blue nail polish is seen interacting with a small figurine of a girl with a purple dress and a blue cape. The hand is holding the figurine over a purple jar with a silver lid. The jar is placed on a white surface. In the background, there is another jar with a blue lid. The style of the video is casual and playful, with a focus on the interaction between the hand and the figurine. The colors in the video are vibrant, with the blue of the nail polish and the purple of the jar standing out against the white surface. The figurine is the main object of interest, with its detailed design and bright colors. The jars in the background add a sense of depth to the scene. The overall impression is one of fun and creativity, as the hand seems to be playfully interacting with the figurine."
+    ],
+    "video_ids": [
+        "BH2PWI-NVkQ_18_0to210"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A hand with blue nail polish, a small figurine of a girl with a purple dress and a blue cape, and two jars (purple and blue).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a hand with blue nail polish interacting with a small figurine of a girl wearing a purple dress and blue cape, positioned on top of a purple jar. A second blue jar is also visible beside it. All core elements described in the condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a hand with blue nail polish, with a small figurine of a girl wearing a purple dress and a cape. which on top of a purple jar labeled Additionally blue jar jar is also visible in the. The elements elements of in the question are present and accurately depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cb5dbf245a776c2b0ea13f72040e426e9b62fce3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cea80d0094976fa7d2dc5084b4c986f50d5813531fe9edc06be88fcd962f936d
+size 122246
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..e2a6f0ed9cfe6842d4ca5c25b21886299b4abff7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a drone shot of a river flowing through a valley surrounded by hills and trees. The river is calm and wide, with a few small rapids visible. The surrounding landscape is lush and green, with a mix of deciduous and evergreen trees. The drone flies over the river, providing a bird's eye view of the landscape. The lighting in the video is soft and diffused, suggesting it may be early morning or late afternoon. The overall style of the video is serene and naturalistic, capturing the beauty of the river and its surroundings."
+    ],
+    "video_ids": [
+        "3WWnYwz8DMw_18_0to185"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: River, hills, trees? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a river flowing through a valley, surrounded by hills and trees. The landscape features natural elements as described, with the river being the central focus, flanked by wooded hillsides. The presence of a road and some distant buildings does not contradict the core description, as they are secondary elements in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a river flowing through a landscape, with by hills and trees. The presence features a elements that described, with the river cutting the central focus, flanked by green hills and. The presence of trees few and a structures structures does not contradict the core description of as they are not elements that the scene.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b65ab7e03afdeea41ad4dee75c65e5c4f90220b3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b024760b17f57a5c7b37f1092f2734e0e27f9cacd13a54b2da5c902e5f625cf1
+size 125445
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..75944b27cc3f1d6022451aec2c805adbfe809cbb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the close-up view of a BMW car's wheel in motion. The wheel is silver with a distinctive BMW logo in the center. The car is red and the wheel is in focus, with the background blurred, emphasizing the wheel's movement. The style of the video is dynamic and action-oriented, showcasing the car's performance and design."
+    ],
+    "video_ids": [
+        "pr4kzOQx1ZE_62_0to116"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: BMW car's wheel, red car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a close-up of a BMW car's wheel, clearly showing the BMW logo and M Sport branding, along with the red car body visible in the frame. The core description of the BMW car's wheel and red car is accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a close-up of a BMW car's wheel, which showing the BMW logo and the Performance emblem, which with the red body body. in the background. The focus elements of the ' car's wheel and the car is fulfilled represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d45488e2b3ce0cdade56a78e383e95eefe97abba
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:619a4d66e28742c0dd69130eb2915d2cd685fba2e97d9cbe1104ef3fb822dbab
+size 225077
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..3f8b09091c69bdd06f44ce89216d7ab87b19f3ff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a car's trunk being opened, revealing a trunk compartment with a spare tire and a toolbox. The spare tire is mounted on the wall of the trunk, and the toolbox is placed on the floor. The trunk is lined with a black carpet, and the spare tire has a yellow \"GOODYEAR\" logo on it. The toolbox is black with silver accents. The trunk compartment is illuminated by the car's interior light. The video is a time-lapse of the trunk being opened, showing the spare tire and toolbox in detail."
+    ],
+    "video_ids": [
+        "AASalydtibk_47_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Spare tire, toolbox? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a spare tire secured with leather straps inside the trunk of a car, along with a toolbox (a black, riveted metal box). These objects are prominently displayed and match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a spare tire in in a straps in the trunk of a car, which with a toolbox.silver black rectangular rectangulareted metal box) These elements are consistent displayed and match the description of.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4a3305cd8eb0a73e11d57a2156496682d9b26c8f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c2bda33749ad1c9ec6f84cc585a527f89876d432ea5f1e98aba4a16eb5a62d13
+size 177848
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..8eec7d276ddfedb1da3e964d919482d48704ac03
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is a quarterback, number 18, who is in the process of throwing a pass. He is wearing a white jersey with blue and orange accents, and his arm is extended, indicating the action of the throw. The quarterback is surrounded by his teammates, who are also dressed in white jerseys with blue and orange accents. They are in various positions, some standing and others in motion, suggesting a high level of activity on the field. The background is filled with the crowd, indicating that the game is taking place in a stadium. The crowd appears to be engaged in the game, adding to the overall atmosphere of excitement and anticipation. The style of the video is realistic, capturing the intensity and action of the game in a way that is both engaging and immersive."
+    ],
+    "video_ids": [
+        "LTkt3hLfN6o_39_150to300"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Quarterback (number 18), teammates, crowd? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing the quarterback (number 18) in a dynamic pose, surrounded by his teammates in formation, and with a blurred crowd in the background. All specified elements are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a quarterback wearingnumber 18) in the white pose, likely by teammates teammates in a. and with a crowd crowd in the background, The elements elements are present and match with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f823e7afefb269f2fe1fe12f0906b1fd395330e7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:715882620ddc2971bb724fc66b757f31b98233b045624999d7f0c078f053a3a8
+size 279586
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..1025fd46001dc1f1d6de7ebc989501435feebd51
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a man in a suit and tie, sitting in front of a backdrop that includes a red banner and a blurred image of a building. The man appears to be a news anchor or reporter, and he is speaking to the camera. The style of the video is professional and polished, with a focus on delivering news content. The man is dressed in a dark suit and tie, and he is seated in a chair that is positioned in front of the backdrop. The backdrop includes a red banner with the words \"BREAKING NEWS\" in white letters, and there is also a blurred image of a building in the background. The man is speaking to the camera, and his expression is serious. The overall tone of the video is serious and informative."
+    ],
+    "video_ids": [
+        "E0vyg8t3qFs_11_21to159"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a dark suit and tie, a red banner with 'BREAKING NEWS' written in white letters, and a blurred image of a building.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a dark suit and tie, a red banner with 'BREAKING NEWS' in white letters, and a blurred image of a building in the background. These elements are all present and match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a dark suit and tie, which red banner with 'BREAKING NEWS' written white letters, and a blurred image of a building in the background. These elements match consistent present and match the description provided.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3b0b752ab0db32cd57d19a1d72cbde23b0094cd3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88ec0310069ce7b1048bd1ed8e4f57f21c57760f6c68a5ce58e6e53463c8216a
+size 122187
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..21914dc735f646c58d5ac66a16013f2058a6625e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two men are seen in a professional kitchen, engaging in a conversation while enjoying a meal. The man on the left, wearing a blue apron, is holding a plate of food and appears to be in the process of eating. The man on the right, dressed in a blue shirt, is holding a fork and seems to be listening attentively to the other man. The kitchen is well-equipped with various appliances such as an oven, a sink, and a refrigerator. The overall atmosphere of the video suggests a casual and friendly interaction between the two men, possibly discussing the food they are eating or sharing their thoughts on the culinary experience."
+    ],
+    "video_ids": [
+        "C9rVBzHvn8M_104_0to197"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, one wearing a blue apron and holding a plate, the other in a blue shirt holding a fork.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men in a kitchen setting. One man is wearing a blue apron over a blue shirt and is holding a plate. The other man is wearing a blue shirt and is holding a fork. Both are interacting with the food on the plate, which matches the description. There are no elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men in a kitchen setting. One man is wearing a blue apron and a white shirt and is holding a plate with The other man is wearing a blue shirt and is holding a fork. The men standing with food food on the plate, which align the description provided The are no additional in contradict the description description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..da3e56ca3cb6c38bb1c2b4d3b43deda06754cd28
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1313-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7541ff619f28debb6d8773831087ef8012dbecc9f24a51f08e713c1cfc02a2b8
+size 180364
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..27ec3de9508e634fe8310ef0f68baf040609375b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the beauty of autumn leaves in a yard. The leaves, in various shades of orange, yellow, and brown, are scattered across the ground, creating a vibrant and colorful display. The leaves are in motion, falling gently from the trees, and covering the ground completely. The yard is surrounded by a white house, which provides a stark contrast to the colorful leaves. The video is shot from a low angle, looking up at the leaves, which adds a sense of depth and perspective to the scene. The overall style of the video is natural and serene, capturing the essence of autumn in a simple yet beautiful way."
+    ],
+    "video_ids": [
+        "6qmiXB5rLGk_8_0to197"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Autumn leaves, a white house? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features autumn leaves covering the ground, which matches the 'Autumn leaves' condition. In the background, a white house is visible, fulfilling the 'a white house' condition. Additional elements like grass and fence do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features autumn leaves in the ground, which align the 'Autumn leaves' part. Additionally the background, there white house is visible, fulfilling the 'a white house' condition. The elements like the and a are not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fb5eb6caf3734bf8b0b35494a58ad0af39e8379e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0b88c1ad4b67d850f479fc2cad089d34f6a31b7047a6de48cf04b5b58d9a7f7
+size 256720
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..0aaa383ab7fe5a9e60444981a107e654d58f897c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a blue and white Smart car parked in a parking lot. The car is positioned in the center of the frame, with its rear facing the camera. The car's license plate is clearly visible. In the background, there are other cars parked in the lot, and a tree can be seen in the distance. The lighting in the video is natural, suggesting that it was taken during the day. The style of the video is straightforward and documentary, with no additional effects or filters applied. The focus is solely on the Smart car, and there is no action or movement in the video. The video is likely intended to showcase the car's design and color scheme."
+    ],
+    "video_ids": [
+        "2VuyezXCvFM_34_0to101"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue and white Smart car, other cars, a tree? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a blue and white Smart car as the main subject, parked in a lot with other cars visible in the background. Trees are also present in the background, matching the described elements. There are no contradictions with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a blue and white Smart car, the main subject. which in a parking. other cars visible in the background. There are also present in the background, fulfilling the description elements. The are no contradictions with the given description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4ed5754e33ba99495e64b3cb0e51cfafb03d9510
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d9011f87825de3ab318036729e3efccab7766f4990c5d105a207776226f3bd48
+size 193493
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..2546744121c10b85822fada94b10b40c82b0a38a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up shot of a Mercedes-Benz steering wheel, showcasing the brand's logo and the airbag. The style of the video is a high-resolution, detailed, and focused on the steering wheel, highlighting the craftsmanship and design of the vehicle. The steering wheel is the main object in the video, and it is captured from different angles and perspectives, emphasizing the Mercedes-Benz logo and the airbag. The video is likely used for promotional purposes, showcasing the luxury and quality of the Mercedes-Benz brand."
+    ],
+    "video_ids": [
+        "A81TxQfaro0_9_0to152"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Mercedes-Benz steering wheel, brand's logo, airbag? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a Mercedes-Benz steering wheel with the brand's iconic three-pointed star logo prominently displayed in the center. The word 'AIRBAG' is visible below the logo, confirming the presence of an airbag. The overall focus and content of the video align with the described objects without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a Mercedes-Benz steering wheel with the brand's logo logo-pointed star logo prominently displayed in the center. The presence 'AIRBAG' is also, the logo, indicating the presence of an airbag. The video composition is details of the video align with the description objects, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1ce340a0c27d1613558ebeb87d60072e9d9ff98a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3abfacaecfcaeaa9c26015c063214a97d6327e8e5b91b005cbeb7f8fa4c39cf7
+size 240627
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..846528b7929e8b6ad53a31d14ba2efb663f4ae39
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up of a black bear in a forest setting. The bear is seen in three different positions, moving from left to right across the frame. In the first frame, the bear is seen peeking out from behind a bush, its eyes focused on something in the distance. In the second frame, the bear has moved slightly to the right, its gaze still directed forward. In the third frame, the bear has moved even further to the right, its head now fully visible as it continues to observe its surroundings. The forest around the bear is lush and green, with various plants and trees filling the background. The bear's fur is a deep black, contrasting with the vibrant colors of the forest. The video is shot in a realistic style, capturing the bear in its natural habitat with great detail and accuracy."
+    ],
+    "video_ids": [
+        "SSKrqaTRngE_36_0to195"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black bear? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a black bear in a forest environment, matching the description. The bear's features, such as its dark fur, facial structure, and surroundings, are consistent with a black bear. There are no elements that contradict the core description of the object being a black bear.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a black bear in a naturaled, which the description of The bear is black, such as its fur fur and sn structure, and the, are consistent with the black bear. There are no additional in contradict the description description of the '.\"\n a black bear.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d98ba8c2ea2885eeabd4b292eb0b84801f333d6f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30c833b34d2dbb629e9920995a8b300f40e72b2404b996764f8a4d0da389d442
+size 201729
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..8c01bd9ba063aee86d30a4ee3fd7e0e71d103850
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a journey through a dense, lush jungle. The viewer's perspective is from the ground, moving forward along a narrow, winding path. The path is flanked by towering trees and thick undergrowth, creating a sense of being enveloped by nature. The ground is uneven and muddy, with fallen branches and leaves scattered about. The viewer's journey is punctuated by the occasional sight of a small stream, its waters flowing gently over rocks and fallen leaves. The overall style of the video is one of exploration and discovery, as the viewer is drawn deeper into the heart of the jungle."
+    ],
+    "video_ids": [
+        "KFtd6MJoUuU_43_0to110"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Narrow, winding path, towering trees, thick undergrowth, fallen branches, leaves, small stream, rocks? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a narrow, winding path through dense jungle, with towering trees, thick undergrowth, and fallen branches visible. A small stream flows through the scene, and rocks can be seen along its banks. All elements described in the condition are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a narrow, winding path surrounded a vegetation vegetation with towering trees and thick undergrowth, fallen fallen branches.. There small stream is alongside the scene, and there are be seen in the path. The the described in the condition are present and contradiction.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4599029345460ef1673370e945458c12bd980edf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59ce5b7ce65ef435468df90b9c05b35610f881e84f1765b329166a4ad2f53421
+size 551340
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..0225eda7c8bbbbe9bf040bb31ee7fdc02d9db51b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in front of a large screen displaying a silhouette of a leopard. The man is wearing a black shirt and glasses, and he appears to be speaking or presenting. The style of the video is a combination of a live event and a visual aid, with the man as the central figure and the leopard silhouette as the backdrop. The focus is on the man and his interaction with the audience, while the leopard silhouette adds a visual element to the presentation. The overall atmosphere of the video is professional and informative."
+    ],
+    "video_ids": [
+        "3wBGeYU1IHM_80_0to101"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man wearing a black shirt and glasses, who is speaking or presenting.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black shirt and glasses, standing in front of a backdrop with a panther silhouette. His mouth is moving, suggesting he is speaking or presenting, which aligns with the 'Object(s)' condition. The background does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black shirt and glasses, standing in front of a background that a leopardther's. Although posture is open, suggesting he is speaking or presenting. which aligns with the descriptionObject(s)' condition. The presence and not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..964e1a194bf8b0939b90eea93b4c9b8015287420
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2a7192c692860733c6856a69e145ddd9d4f259ca1944e32163654b80e5e4e65f
+size 264208
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..4476279800d6da981e0b9a95037106ff2cf9f13e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and mustache, wearing a white shirt and a necklace. He is standing in a room with a window in the background. The man is looking to the side with a serious expression on his face. The room appears to be well-lit, and the window suggests that it might be daytime. The man's attire and the setting suggest that this could be a scene from a movie or a television show. The overall style of the video is realistic, with a focus on the man's facial expression and the room's ambiance."
+    ],
+    "video_ids": [
+        "KTvdiIWenk8_5_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and mustache, wearing a white shirt and a necklace.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The man in the video has facial hair (beard and mustache), is wearing a white shirt, and a black necklace. These elements match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video in the video has a hair thatbeard and mustache), is wearing a white shirt, and has necklace necklace. The elements match the description provided. the questionObject(s)' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a4b98b9e7d12520fb244b5931c7e1e4d91d67620
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb4c745e3da1c08df56fff197ae96e2dc9b75a78886cfcf97cfceb9b48eb0efd
+size 128065
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..96ad9e7284125793ebfb725831772eb734416b10
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a young boy with curly hair and glasses, dressed in a formal outfit with a black vest and an orange tie. He is smiling and appears to be in a car showroom or garage, as there is a red car in the background. The boy is looking up and to the right, possibly at something or someone off-camera. The style of the video is candid and informal, capturing a moment of joy and excitement. The boy's outfit and the setting suggest that he might be attending a special event or occasion. The video does not contain any text or additional graphics."
+    ],
+    "video_ids": [
+        "cBMH7B9LEx4_5_0to186"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young boy with curly hair and glasses, wearing a formal outfit including a black vest and an orange tie, smiling and looking up and to the right.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young boy with curly hair and glasses, wearing a formal outfit including a black vest and an orange tie. He is smiling and looking up and to the right, matching the description. The presence of a red car and another person in the background does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a young boy with curly hair and glasses, wearing a formal outfit including a black vest and an orange tie. He is smiling and looking up and to the right. which the description provided The background of a red car in a person in the background does not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2277b0f132d0f4d8c7c1d3662b9a5678669adf85
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1314-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9bf95dad74bcf53fe936d1d18aa838c20f5b1d080bb6ee749b95bc4a784e3ce4
+size 138764
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..d84db53c8c40e293d9fc00b7b9cf6141abd8a2ff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a small black puppy sitting inside a large black tire. The tire is placed on a grassy field. The puppy appears to be curious and is looking around the tire. The puppy is the main subject of the video, and the tire is a significant object in the scene. The grassy field provides a natural and open setting for the puppy's exploration. The video captures a moment of the puppy's life, showcasing its curiosity and playfulness."
+    ],
+    "video_ids": [
+        "u3JGSo_pyXc_28_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small black puppy? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a small black puppy sitting inside a tire on grass, which matches the description of 'A small black puppy'. There are no conflicting elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a small black puppy inside inside a tire, a. which matches the description of aA small black puppy'. The are no additional elements in contradict this description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a7f817f464922b6b1336a644471adec4e2c69bf7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0a4a4a1b0104ee2b508452b8f727ed1c5d74b15f9f54d28bda6db237d1e7b1f
+size 284857
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..f1397f96c4f5af20f4a7efe933d16cb30b13ce6c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen sitting at a dining table, holding a fork and a knife. They are in the process of cutting into a blue plate filled with food, which includes a piece of meat, potatoes, and a side of cranberry sauce. The person is wearing a plaid shirt and appears to be enjoying their meal. The table is set with a blue plate, a fork, and a knife. The scene is set in a casual dining setting, with the person enjoying their meal in a relaxed manner. The focus of the video is on the person and their interaction with the food on the plate."
+    ],
+    "video_ids": [
+        "8wBJLWsVWI8_46_0to110"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a blue plate, a fork, a knife, a piece of meat, potatoes, and cranberry sauce.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person eating from a blue plate, using a fork and knife. The plate contains what appears to be a piece of meat (possibly sausage or bacon), potatoes, and cranberry sauce. The scene is consistent with the described objects, even though the person is partially obscured and the cranberry sauce is on a pancake or similar item, not directly on meat. No elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person using a a blue plate. which a fork and knife. The plate contains a appears to be a piece of meat,possibly a), a), potatoes, and aberry sauce. The presence is consistent with the description objects, and though the person's not out and the focusberry sauce is not the plateake rather similar food, which on on the as The additional contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f173c77a7964aa40268b8af1423e0c497c586e9d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:162a09131b65a734ffeddf58237a48b71d6b5c9eacea3ae9664e05fbbf725b8d
+size 186687
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..951e20c4325d46b2728bd22c0338041c21feb899
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a chef is seen preparing a dish in a professional kitchen. The chef, wearing a white coat and gloves, is meticulously measuring and mixing ingredients. The chef is using a silver bowl and a measuring cup to combine the ingredients. On the counter, there are several bottles, possibly containing oils or sauces, and a bag of flour. The chef's focus and precision suggest a high level of skill and attention to detail. The overall style of the video is clean and professional, emphasizing the chef's technique and the quality of the ingredients."
+    ],
+    "video_ids": [
+        "yL-SGBHhHNk_4_21to214"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chef, silver bowl, measuring cup, bottles, bag of flour? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person in a chef's coat and gloves, consistent with a chef. A silver bowl, measuring cup, two small bottles, and a bag of flour are all visible on the table. The actions align with the described objects being used in a preparation process.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person in a chef's uniform and gloves mixing which with the chef, The silver bowl is bottles cup, and bottles bottles, and a bag of flour are visible present on the counter, The person performed with the chef scenario, used in a cooking process,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4b1966607eee71a2a8ed33424a763061d7bfebaa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8441a6fcd39ebd3ba9d9ff92443fd00e944c71553bb69b3226df35b1e751801c
+size 135802
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..fee8d956a245c527f0d318ccc2d2000fe0df246b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment from a football game. The main focus is on a football player, who is wearing a white jersey with the number 2 and the name \"MANZIEL\" on the back. He is holding a football in his right hand and is walking away from the camera. The player is wearing a helmet and is wearing white pants. The player is walking on a football field. In the background, there is a referee who is wearing a black and white striped shirt. The referee is holding his arms out to the sides. The referee is standing on the football field. The style of the video is a sports video. The video captures the action of the game and the players. The video is shot in a way that the player is the main focus and the referee is in the background. The video is shot in a way that the player is walking away from the camera. The video is shot in a way that the player is holding a football. The video is shot in a way that the player is wearing a helmet. The video is shot in a way that the player is wearing white pants. The video is shot in a way that the player is wearing a white jersey. The video is shot in a way that the player is wearing a white shirt. The video is shot in a way that the player is wearing a white helmet. The video is shot in a way that the player is wearing a white jersey with the number 2 on it. The video"
+    ],
+    "video_ids": [
+        "BXS5VJnXHN8_8_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: football player, football, helmet, white pants, white jersey, white shirt, white helmet, referee? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a football player (Manziel, #2) wearing a white jersey, white pants, and a white helmet with orange accents. He is holding a football. A referee in a black-and-white striped shirt is also visible, raising his hand. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a football player wearingnumberziel) wearing2) in a white jersey, white pants, and a white helmet. a details. The is holding a football. There referee in a black and-white striped shirt is also visible in fulfilling his hand, The the elements of in present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..81a4c3d1382fe64ef812140e9576d1f0ce4bb95a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7d0d1a469b3abae92e6a8a747f6263433f62be21d809446e77958cccfbedd19c
+size 199887
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..55256b5a12720c74fc495eb6f549913216fd682b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a vibrant and colorful display of a healthy smoothie being prepared. The first frame shows a glass of red smoothie with a few fresh ingredients sprinkled on top, including a sprig of dill and a slice of apple. The second frame shows the same glass of smoothie, but now with a few more ingredients added, including a slice of beet and a sprig of parsley. The third frame shows the final product, a fully prepared and garnished glass of red smoothie, ready to be enjoyed. The style of the video is simple and straightforward, focusing on the preparation of the smoothie and the fresh ingredients used. The colors are bright and appealing, emphasizing the healthiness and freshness of the ingredients. The video is likely intended to inspire viewers to make their own healthy smoothies at home."
+    ],
+    "video_ids": [
+        "ZwryouK_dfk_46_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ['glass of red smoothie', 'sprig of dill', 'slice of apple', 'slice of beet', 'sprig of parsley']? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a glass of red smoothie topped with a sprig of dill, a slice of apple, and slices of beet. There are also sprigs of what appears to be parsley or a similar herb scattered around the scene. All specified objects are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a glass of red smoothie, with a sprig of parsleyill. which slice of apple, and a of beet. The is also additionaligs of parsley appears to be parsley in a similar green garn around the glasses. The the objects are present, the with the description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..53435cbc80db0c0f2e320ff1cc2666e7abb75d1b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:023cb02075927e5d199cc243da12718cd3b011f73dfb40a4ae1a87a773821685
+size 127024
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..58b0cbe4db3386b2e7be6cba361988a694b9150d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a black and red motorcycle wheel with a black tire and red rim. The wheel is positioned on a gravel road, and the motorcycle is stationary. The background features a rocky cliff with sparse vegetation, and the sky is clear and blue. The style of the video is a straightforward, static shot with no movement or action, focusing on the details of the motorcycle wheel."
+    ],
+    "video_ids": [
+        "cMHezRztzR4_12_0to150"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black and red motorcycle wheel with a black tire and red rim.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a scooter wheel, which has a black tire and a red rim, matching the description. The wheel is attached to a black frame, and the overall appearance aligns with the 'black and red motorcycle wheel' description, even though it's a scooter. The presence of additional elements like the gravel ground and background does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a motorcycle wheel that which is a black tire and a red rim, matching the description provided The presence is part to a black body, and the background appearance aligns with the givenblack and red motorcycle wheel' part. even though it is a scooter wheel The presence of a elements like the stone and and the rock not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cc21c5595c8000b89cf9cf5099035e020af13e73
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2fce55044aabae1365578511746a8e53ef1b5b1cd70115419bac722dc1bc0146
+size 151655
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..301a5dc23b9b6b2b3108a35ca98763ef7ad3f5c2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a serene rural scene with a brown horse standing in a lush green field. The horse is positioned in the center of the frame, looking directly at the camera. In the background, there is a rustic wooden barn with a sloping roof, surrounded by a fence. The barn is situated on a hill, and the sky above is clear and blue. The overall style of the video is calm and peaceful, capturing the beauty of rural life."
+    ],
+    "video_ids": [
+        "3wxqBMV4C8M_423_97to228"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A brown horse? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a brown horse standing in a field in front of a wooden barn. The horse is clearly visible and matches the description provided. Additional elements such as people, other animals, and farm equipment do not contradict the core description of a brown horse.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a brown horse in in a grass, front of a barn barn. The horse is the the and matches the description of. The elements such as the in a animals, or the equipment are not contradict the core description of the brown horse in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..eb3d1f8ec81cf86aa7d0353634c961617fa5f080
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:749077d9dcb9eef70fb52841ae1555de039ec1bec332a00b14bd6f1eb5499729
+size 222306
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..1fad2417f44831246284c2eb594e33c485476345
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a black shirt, who appears to be engaged in a conversation. The man is looking to his left, suggesting that he is interacting with someone off-camera. The background is dark, which puts the focus on the man. The style of the video is a close-up shot, which emphasizes the man's facial expressions and body language. The lighting is dim, creating a moody atmosphere. The man's expression is serious, indicating that the conversation might be of importance. The overall style of the video suggests a dramatic or intense scene."
+    ],
+    "video_ids": [
+        "tmFqmYtDHkE_15_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a black shirt, which matches the 'Object(s)' condition. The background is dark, and the focus is on the man, with no conflicting elements present that would contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a man wearing a black shirt. which matches the descriptionObject(s)' condition. The presence and not, and the man is on the man, making no additional elements..\"\n would contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..76dfe90de0deab984d200da4a871008bb81f3448
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1315-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9cf1bf6e743c1d33a0c80a1d9130bcd962b429883dee2518227dc2b3b103ac9e
+size 147426
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..f892aceded82e3c3661543ae2d49243b8e5941b5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a group of Buddhist monks in a serene setting. They are dressed in traditional yellow and red robes, signifying their religious affiliation. The monks are seated on the floor, engaged in a peaceful activity. In front of them, there is a table with a stack of books, suggesting that they are studying or meditating on the teachings of their faith. The overall atmosphere of the video is calm and contemplative, reflecting the spiritual nature of the monks' activities. The video captures the essence of Buddhist practice, emphasizing the importance of study, meditation, and community in the pursuit of enlightenment."
+    ],
+    "video_ids": [
+        "1DDWMwG42iM_36_0to102"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A group of Buddhist monks seated on the floor, a table with a stack of books.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group of Buddhist monks seated on the floor, dressed in traditional orange and red robes, engaged in study or prayer. A small table with a stack of books and scrolls is visible in the foreground, matching the described elements. Additional details like monks handling books or prayer beads do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a group of individuals monks seated on the floor, which in traditional yellow robes black robes, which in a or meditation. In table table in a stack of books is a is placed in the foreground, which the description elements. The elements like the' the and the beads are not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d105b5ff5466dbbbaf37588dd1c26d733ceeaec6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb1e283de004e3dfe92423906d0b9c840218c15ccd698daa16726502a8b3b8db
+size 127265
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..629ce490805340f6a4496eac9535f8f3b84675d0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a kitchen, preparing a cake. She is wearing a blue shirt with the words \"How to Cake It\" written on it. The kitchen is equipped with a refrigerator and a shelf filled with various kitchen utensils and ingredients. The woman is seen pouring white frosting onto a cake, which is placed on a wooden table. The cake is white and appears to be a small, round cake. The woman is focused on her task, and the kitchen is well-lit, creating a warm and inviting atmosphere. The video captures the process of cake decoration, showcasing the woman's skill and attention to detail."
+    ],
+    "video_ids": [
+        "2ShTNbxSZnU_94_0to134"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a white cake on a wooden table, white frosting, a blue shirt with 'How to Cake It' written on it.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a blue shirt with 'How to Cake It' written on it, standing at a wooden table with a white cake shaped like an igloo. She is sifting white frosting (or powdered sugar) over the cake. All core elements from the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a blue shirt with theHow to Cake It' written on it, standing at a wooden table. a white cake on like a iceloo. She is pouringifting white frosting ontolikely a sugar) onto the cake. The the elements of the description are present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..94e02cafeebecd4b53244536a3d0a52bf893cdac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e6c5b51331213253385b88182c84dcad72f573813c8178a9b0fdf7d87873996
+size 118287
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..c90396c600f37f6b416c64158d9de587759eb68a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with black hair and glasses, smiling and gesturing with her hands. She is wearing a dark blue dress with white stripes. The background is a colorful cityscape with buildings and lights. The style of the video is a television interview or news segment, with a focus on the woman's expression and body language."
+    ],
+    "video_ids": [
+        "0am2J7c25Rc_3_0to181"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with black hair and glasses, smiling and gesturing with her hands, wearing a dark blue dress with white stripes.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with black hair and glasses, smiling and gesturing with her hands. She is wearing a dark blue dress with white stripes, which matches the description. The background and other elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with black hair and glasses, who and gesturing with her hands. She is wearing a dark blue dress with white stripes. which matches the description. The background, additional elements are not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f197b728a41aa6861ee63171c4ceebad35b444cc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2abb229a715eab6404f1345f3a37870d142c2fb6b0837678f88fe12a5e554e52
+size 208721
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..a80b7ee89df617a439d70c1915ea9d2a8056c504
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen cooking in a kitchen. The person is using an orange pot on a gas stove. They are stirring the contents of the pot with a spoon. The kitchen is well-equipped with various utensils and ingredients. The person is focused on their task, ensuring that the food is cooked evenly. The overall atmosphere of the video is one of a homey and comfortable kitchen environment."
+    ],
+    "video_ids": [
+        "IqS9HuXm9iQ_42_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, an orange pot, a gas stove, a spoon? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person covering an orange pot on a gas stove. Although a spoon is not clearly visible in the frames, the presence of a wooden spoon on a tray nearby suggests it is part of the kitchen scene. The core elements (person, orange pot, gas stove) are clearly present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person cooking an orange pot on a gas stove with The the spoon is not explicitly visible in the video provided the person of a spoon spoon in the plate in suggests that could part of the cooking setup. The person elements ofa, orange pot, gas stove) are present depicted and match the description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..56b66bef68a47c0f13105530a29b9f90672e6461
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7adbd5162a3e4dac41d7193f26e67b528719d865ebca8ba4d03a653cf3a94ace
+size 194187
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..e092011eaa694d23d93459ffd329e562600c7972
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red Jeep parked on a dirt road. The Jeep is equipped with large off-road tires and a black front bumper. The vehicle is positioned in front of a tree, and the sun is shining on the hood, creating a reflection. The Jeep's grill and headlights are clearly visible. The style of the video is a straightforward, real-life depiction of the vehicle, with no additional embellishments or artistic effects. The focus is on the Jeep and its immediate surroundings, with no other significant objects or actions in the scene."
+    ],
+    "video_ids": [
+        "wVPtEg3zSjU_14_0to160"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red Jeep, tree? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red Jeep, which is the main subject, and a tree is clearly visible in the background to the left of the vehicle. Both elements are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a red Jeep, which is the main object. and there tree is visible visible in the background. the left of the vehicle. The elements are present and match with the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ddb99c75d03911a126a35f2dd8d36b9aaa03bedc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c39add45a60b52316ba0fe24f2839f9b503e566d282275319856d503fed5e83
+size 227026
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..7a758f2f963b9543a791b6e24e3dab2a6b1ab7aa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a step-by-step tutorial on how to make a 6-layered chocolate mousse cake. The style of the video is a close-up, high-resolution food photography with a focus on the cake and its ingredients. The cake is shown in various stages of preparation, from the initial layers to the final decoration. The ingredients include chocolate, cream, and various chocolate-covered items such as cookies and candies. The video is likely to be informative and visually appealing, with a focus on the textures and colors of the cake and its components. The overall tone of the video is likely to be warm and inviting, encouraging viewers to try making the cake themselves."
+    ],
+    "video_ids": [
+        "hXDwzE2bvBo_0_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: 6-layered chocolate mousse cake, chocolate, cream, chocolate-covered cookies, chocolate-covered candies? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a 6-layered chocolate mousse cake with a glossy chocolate glaze. It is decorated with chocolate-covered candies (white and brown spheres) and chocolate-covered cookies (the crescent-shaped pieces). The cake also includes chocolate-striped wafer sticks, which are chocolate-based. All elements described in the condition are present, and no conflicting elements are shown.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a cake6-layered chocolate mousse cake with alternating glossy chocolate coatingaze. The is adorned with chocolate-covered cookies andpossibly and dark)) and chocolate-covered cookies (ch rectangularcent-shaped pieces on The presence is has cream-covered layersfer cookies, which are a-covered. The these mentioned in the question are present and and the conflicting elements are introduced.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8544d057302579cb8016351973b2af3fe599d81a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b1032c150c7460be73cf1c024a0b76432d3afd7122dc901fdb2d3b9625509f6
+size 107930
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..fa7610eb6545b63e52767c0ce46049bb0669b21c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up, high-resolution, and detailed view of the engine bay of a BMW TwinPower Turbo vehicle. The style of the video is informative and technical, showcasing the intricate components and design of the engine. The engine cover prominently displays the BMW logo, and the TwinPower Turbo text indicates the type of engine. The video likely provides a comprehensive view of the engine's components, such as the air intake system, battery, and other mechanical parts. The focus is on the engine's design and functionality, rather than the aesthetics of the vehicle."
+    ],
+    "video_ids": [
+        "6prSqgUIgQU_20_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Engine cover, air intake system, battery, mechanical parts? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the engine cover with the BMW logo and 'TwinPower Turbo' text, the air intake system with its ducts and housing, a visible battery with a red cap, and various mechanical parts such as hoses, connectors, and engine components. All specified elements are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a engine cover of the BMW logo, surroundingMWINPower'' text, which air intake system with visible metallics and vents, and glimpse battery with a blue terminal, and various mechanical parts such as the and cl, and the components. The these objects are present and match depicted,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..42da2e928838d96a9c5362329a6f907c08bff9a8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c726457f3de9abfc57ef8a5fc78b93ee872a3e320b6ae742165f286a1c2fa1d6
+size 110393
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..eac4af40b22f138cc788ba8c32437894e0781eaf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene moment of an orange tabby cat perched on a green wooden fence. The cat, with its fur glowing in the sunlight, gazes intently to the left, its eyes reflecting curiosity and alertness. The fence, painted a vibrant green, provides a stark contrast to the cat's orange fur. In the background, lush green foliage adds a touch of nature to the scene. The cat's position on the fence and its focused gaze suggest it might be observing something of interest in the distance. The overall atmosphere of the video is calm and peaceful, with the cat as the main subject of the viewer's attention."
+    ],
+    "video_ids": [
+        "z_PfMpciox0_10_0to187"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An orange tabby cat? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a clear depiction of an orange tabby cat resting on a light green surface, with its characteristic orange fur and striped pattern visible. The cat's features, including its ears, eyes, whiskers, and paws, are consistent with the description of an orange tabby cat. There are no elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a cat depiction of an orange tabby cat sitting on a green green wooden, likely a fur orange fur and white pattern.. The cat's posture, such its whisk, whisk, anders, and theaws, are consistent with the description of an orange tabby.. The are no additional in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..86c648ba27870f4fcc236947a832e099a849750b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1316-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2277cb46dc1e25b8a28ab74246915d0eef360955b1f9e7dd7f965eab5b52b38f
+size 263543
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..140c90855c4019368b3c657f6168f636fd133c12
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a white plate filled with fresh fruit, including slices of pink grapefruit and blackberries. The fruit is garnished with green leaves and purple flowers, adding a touch of color and elegance to the presentation. The plate is placed on a white surface, which contrasts with the vibrant colors of the fruit. The camera angle is slightly above the plate, providing a clear view of the fruit and its garnishes. The focus is on the fruit, with the background blurred, drawing attention to the main subject. The overall style of the video is simple yet elegant, highlighting the natural beauty of the fruit and its presentation."
+    ],
+    "video_ids": [
+        "nZ4j7OfHaGI_1_135to331"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: White plate, slices of pink grapefruit, blackberries, green leaves, purple flowers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a white plate with slices of pink grapefruit, blackberries, green leaves (mint), and purple flowers (herb sprigs). All specified elements are present and accurately represented in the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as showcasing showing a white plate with slices of pink grapefruit, blackberries, green leaves,likely or and purple flowers.possiblyb).ig). The the elements are present and match depicted, the image content.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..851e62813bd41775c2087f1ddbe55deb15fd0692
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:158998ab64fdcd39ba13d259ac7f633784b39622f5d90d09b539de9f11cde518
+size 95137
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..36791116a317e87802f7a6475a0d3eedea9d4385
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man in a yellow safety vest is seen loading a red, rusted car onto a flatbed trailer. The car is parked on a grassy area, and the man is using a winch to pull it onto the trailer. The trailer is attached to a blue truck, which is parked nearby. In the background, there are other vehicles, including a white truck and a green car. The scene takes place in a wooded area, with trees visible in the background. The man appears to be working alone, and the overall atmosphere of the video is one of manual labor and outdoor activity."
+    ],
+    "video_ids": [
+        "asuB6I1r7a0_9_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a yellow safety vest, a red, rusted car, a flatbed trailer, a blue truck, a white truck, and a green car.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a man in a yellow safety vest working near a red, rusted car that is being loaded onto a flatbed trailer. A blue truck (the tow truck) and a white truck are visible in the background, along with a green car further in the distance. All core objects mentioned in the description are present and accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts ful a man in a yellow safety vest, on a red, rusted car on is on loaded onto a flatbed trailer. The blue truck isthe flat truck) is a white truck are visible in the background. and with a green car. in the distance. The the elements are in the description are present and correctly represented in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..26beecd6fc05206d96c9f52abeb1893dbc530d0d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:25efa8f2a9bddc46bc971d6c182b9e9166acac9865f364f5aa6f52ccbbac2866
+size 160281
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..83bc52af004b0847afeb6a24758eae98303b37fa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a lively scene at a movie premiere. A group of four people, two women and two men, are standing on a stage. The man in the center is holding a microphone, suggesting he is the host or a speaker. The backdrop is a vibrant blue, adorned with white geometric shapes, adding a modern touch to the event. The attendees are dressed in stylish outfits, with the women wearing dresses and the men in jackets, indicating a formal occasion. The overall atmosphere is one of excitement and anticipation, as the group seems to be waiting for the event to begin."
+    ],
+    "video_ids": [
+        "ClqkkWtzNO8_9_18to139"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Four people (two women and two men) standing on a stage. A man in the center holding a microphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows five people standing on a stage, which includes two women and three men. The man in the center is holding a microphone, which matches the description. The presence of an additional person does not contradict the core description, as the condition only requires four people (two women and two men), and the additional person is not a contradiction since the condition does not specify an exact number beyond 'four people'.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows four people standing on a stage, with slightly two women and two men. However man in the center is holding a microphone, which ful the description. The presence of an additional person does not contradict the core description as as the main only specifies four people,two women and two men) and the video person does acceptable a contradiction.\"\n the core does not specify the exact number of fourfour'.'.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c2a0184387b6429e6b8d137e61783d10e72d2f70
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7bc7928ba3881c5b6a86810a68314598aa64744d86c7cb284741679dd6951026
+size 164700
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..07d9c5e76428a882a873914aca91db0ad8749add
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a luxury car from the perspective of the passenger seat. The car's interior is predominantly white, with black accents on the door panels and center console. The steering wheel is on the right side of the car, indicating that the car is designed for driving on the left side of the road. The dashboard features a large infotainment screen, and the car is equipped with a sunroof. The car appears to be parked in a lot, as evidenced by the presence of other vehicles in the background. The style of the video is a straightforward, unedited shot of the car's interior, with no additional graphics or text. The focus is solely on the car and its features, providing a clear view of the vehicle's design and layout."
+    ],
+    "video_ids": [
+        "-foJGIEqfyo_13_0to126"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, dashboard with an infotainment screen, sunroof, door panels, center console? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows the steering wheel, dashboard with an infotainment screen, sunroof (visible at the top of the frame), door panels (visible on the left side), and center console (with controls and storage). All core elements described are present and clearly visible.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful the interior wheel, dashboard with an infotainment screen, sunroof,part at the top of the frame), door panels,on on the left side), and center console (visible the and a compartments The these elements of in present and clearly visible in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..637343fd8356fb3c374feaf4bab64dd0d7f348f8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a2d50fdcb3d0796ad4e5b54c8489d195c421276d2c2057698bb154f140004ec9
+size 225236
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..6611e09e1521d3ef93243a1219acc944f61d2e75
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a kitchen, speaking to the camera. He is wearing a white and blue t-shirt. The kitchen is well-equipped with white cabinets and a wooden countertop. On the countertop, there is a microwave, a toaster, and a coffee maker. The man appears to be in the middle of a conversation, possibly sharing a recipe or cooking tip. The overall style of the video is casual and informative, with a focus on the man and his surroundings."
+    ],
+    "video_ids": [
+        "1OEoJS38q1Q_5_64to221"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a white and blue t-shirt, a microwave, a toaster, and a coffee maker.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a white and blue t-shirt, and in the background, there is a microwave and a toaster on the kitchen counter. While a coffee maker is not explicitly visible, the presence of other kitchen appliances and the overall context suggest a kitchen setting where a coffee maker might reasonably be expected. The core elements described are present, and additional elements do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man wearing a white and blue t-shirt. standing in the background, there is a microwave, a toaster on the counter counter. Additionally there coffee maker is not explicitly mentioned, the presence of a kitchen appliances like the overall setting suggest that kitchen setting where a coffee maker could be be located. The video elements of in present, fulfilling the elements do not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cb37bb0ab59b7c1186fcfff86486d4e255d4576e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24b5101246937c8585a98cea8848351cfdb0c0faebdad97f751cdfbc87e0ddca
+size 118604
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..313d70d63d74dfe45366f536a1ffd710f07aa254
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young man with glasses, wearing a green shirt, standing in front of a shelf filled with various toys and collectibles. The man appears to be speaking or gesturing with his hands, suggesting he might be explaining something or engaging in a conversation. The toys on the shelf include a variety of characters from different franchises, such as Disney, Pixar, and possibly others. The shelf is well-organized, with each toy placed neatly next to the others. The room has a casual and comfortable atmosphere, with the man and the toys being the main focus of the video. The style of the video seems to be a casual, informal discussion or presentation, possibly related to the toys or collectibles on the shelf."
+    ],
+    "video_ids": [
+        "2o84E4vLo8A_6_304to497"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man with glasses, wearing a green shirt, and a shelf filled with toys and collectibles.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a young man with glasses wearing a green shirt, standing in front of a shelf filled with various toys and collectibles. The core elements described are accurately represented, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a young man wearing glasses wearing a green shirt. standing in front of a shelf filled with various toys and collectibles. The description elements of in present represented in and there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..72992566799907f5b3d14575eab328932be78daa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:65fb6c45c6b26487f9ed4a998cbf61538c81ce8a58b75f02b73e68a400d0a38e
+size 199922
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..9920e88abe008f43a8b75b9d36eb6abf1699db9a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a bald man with a beard and glasses, wearing a blue suit and a white shirt. He is seated in a room with a brick wall and a bookshelf in the background. The man appears to be speaking or reacting to something, as he is making a facial expression. The lighting in the room is soft and warm, creating a comfortable atmosphere. The man's glasses and beard give him a scholarly or intellectual appearance. The brick wall and bookshelf suggest a setting that could be a study or a personal library. The overall style of the video is realistic and naturalistic, with a focus on the man's facial expression and the room's ambiance."
+    ],
+    "video_ids": [
+        "7RJrnyB8jnQ_0_0to187"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bald man with a beard and glasses, wearing a blue suit and a white shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man with a beard and glasses, wearing a blue suit and a white shirt, which matches the description. The background and additional elements (like the Twitter handle overlay) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man with a beard and glasses, wearing a blue suit and a white shirt. which matches the description provided The background appears additional elements dolike the brick logo)) do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cc0dbc673b80b1101b0881ea9350418ff211d9e0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5bfba42fd01df19d0ac7f60e855e635a06640ac888a9fc51f12935d7f8bb6fc4
+size 174341
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..0e9e94c05185adb301c1c1b1af57603a5243d37c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a lively scene at a skate park. The park is bustling with activity, with skateboarders performing tricks on the ramps and rails. The ramps and rails are made of concrete and are in various shapes and sizes, providing a variety of challenges for the skateboarders. The park is surrounded by trees, adding a touch of nature to the urban setting. In the background, there's a soccer field, indicating that the park is part of a larger sports complex. The video is shot from an aerial perspective, providing a bird's eye view of the park and its surroundings. The skateboarders are dressed in casual clothing, suitable for skateboarding, and are seen in various stages of their tricks. The video captures the energy and excitement of skateboarding, as well as the beauty of the park and its surroundings."
+    ],
+    "video_ids": [
+        "CvLcQOojE2M_3_226to422"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Skateboarders performing tricks on the ramps and rails.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an aerial view of a skate park with multiple ramps and rails, and numerous individuals who appear to be skateboarders actively performing tricks. The motion and positioning of the people on the skate park structures are consistent with skateboarding activities, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a animated view of a futuristicpark with ramps ramps and rails. and a skateboard are appear to be performingers performing performing tricks. The presence and positioning of the figures suggest the ramps park structures suggest consistent with theboarding activities, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6994bc79481aa4915b02dc87bece117ad065a34c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1317-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3c7d2aa74b8d3aaf27026e3f8238c4c5385522e6cee0e44f575613087ec3efa
+size 205411
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..9d9ebbccc51c34a575754ed63fbf7ead102bc487
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are seen in a forest setting. The man, who is wearing a leather jacket, has a bloody nose and a bruised face, suggesting he has been in a fight. Despite his injuries, he is smiling and looking at the woman with a sense of amusement. The woman, who is wearing a black dress, is looking at the man with a concerned expression. The forest around them is lush and green, with trees and bushes filling the background. The overall style of the video is realistic, with a focus on the interaction between the two characters."
+    ],
+    "video_ids": [
+        "RQn_4Mu8B6U_11_0to161"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman interacting in a forest setting. The man has visible injuries, and the woman is seen from behind, confirming the presence of both individuals as described in the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a man and a woman standing with a close setting. The presence is a facial on and the woman is dressed from the, looking the presence of two individuals as described. the '.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dfc42680c8f1b626da5d391a9becffbce0b88b87
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f0b7c5c7e3530fa035e722bc227309c26fd953f7f21d960dc1fb2271e0621aff
+size 233500
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..827b98699dd498dea4d860de8e805cf4f4a83ea5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a festive scene with a collection of wine glasses and vases. The glasses are elegantly arranged on a table, reflecting the ambient light. The vases, varying in size and shape, are filled with lush greenery, adding a touch of nature to the setting. The table is set against a backdrop of a store, where a display of gold-colored items catches the eye. The overall style of the video is elegant and sophisticated, with a focus on the intricate details of the glassware and the vibrant colors of the vases and greenery. The video is likely to be used for showcasing the glassware and vases in a store or for a special event."
+    ],
+    "video_ids": [
+        "G7BIe9z6xKU_23_0to199"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Wine glasses, vases, greenery, gold-colored items? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows wine glasses (champagne flutes), vases (carafes), greenery (pine branches), and gold-colored items (coasters and trays). All these objects are present and consistent with the description, even though other decorative elements are also visible in the background.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows wine glasses,bothampagne andutes and vases (blueryinges), greenery (plantsapp), and gold-colored items (possiblyasters and possibly) The these elements are present and match with the description, fulfilling though the elements elements like also visible, the background,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f12f458c85cc9dececf01650fc2e34d3ef818cd4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fc900245de3939fe48eb9799991b55f37db3dc628ceb800936778190e8fc74db
+size 242281
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..9e28fd3408931afbc99fb11afcd6aa50ccd26e78
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a vibrant garden scene, showcasing a variety of orange flowers in full bloom. The flowers, with their bright orange hue, are the main focus of the video, drawing the viewer's attention. The garden is lush and green, with a variety of other plants and trees providing a rich backdrop to the orange flowers. The video is shot from a low angle, which emphasizes the height of the flowers and the depth of the garden. The sunlight filters through the leaves, casting a warm glow on the flowers and highlighting their vibrant color. The overall style of the video is naturalistic, capturing the beauty of the garden in a realistic and unfiltered manner."
+    ],
+    "video_ids": [
+        "sJpA0EVaHrU_31_0to135"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bright orange flowers in full bloom? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features bright orange flowers in full bloom, which are the central focus of the scene. The surrounding green foliage and trees do not contradict this description and are consistent with a lush garden setting. The flowers are clearly visible and vibrant, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features bright orange flowers in full bloom, which align the central focus of the scene. The vivid greenery and additional provide not contradict the description but serve consistent with a garden garden setting where The vivid are vivid visible and vibrant, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2edb28404cb5f6904809abad92cd8a87b8d088a9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:791eacfe625bd31708f28698c3f1aaf1a415c7b4a2a1f639e6ce0d29098c624e
+size 220349
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..9e0f44cd6c8a01da32c6a4a5e676f4f9a873c622
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment about a young boy's dream coming true. The boy is standing next to a police officer in front of a helicopter. The helicopter is white with blue and red stripes. The boy is wearing a baseball cap and a jacket. The police officer is wearing a uniform. The background shows a clear sky. The news segment is from ABC 15 in Phoenix. The title of the segment is \"Valley Boy's Dream Come True\". The segment is likely to be about the boy's experience of getting to ride in the helicopter with the police officer."
+    ],
+    "video_ids": [
+        "_4v97awBYWw_6_74to254"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young boy, a police officer, and a helicopter.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a young boy and a police officer standing next to a helicopter. The helicopter is marked with 'STATE TROOPER' and 'Make-A-Wish Arizona,' indicating it's a public safety vehicle. The scene matches the core description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a young boy standing a police officer standing in to a helicopter. The boy is labeled with theV POLROOPER' and 'V a-Wish',', which it is likely special service vehicle. The presence is the description description of any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9c66a292f639bf23411f7d31d2cffdfc230d3dec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d4de8c729d2a016edc0eea48be3edb6dca17bc914840d22b8ebaebc67af1975b
+size 213267
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..e05304da2bf209e1f78b658da9705b77bd256d7b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a blue suit and a pink tie, sitting in front of a white background with a logo in the top left corner. The man appears to be speaking, as suggested by his open mouth and attentive expression. The style of the video is formal and professional, likely a news segment or a corporate interview. The man's attire and the logo in the background suggest that the video is related to a formal event or organization. The white background provides a clean and uncluttered look, allowing the viewer to focus on the man and his message. The logo in the top left corner indicates that the video is likely associated with a specific organization or company."
+    ],
+    "video_ids": [
+        "0yD-o44O2SM_10_27to148"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue suit and a pink tie, speaking.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue suit and a pink tie, who is actively speaking. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue suit and a pink tie, which appears speaking speaking. The setting includes attire elements are not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e499dcfc123342408f575262463ae3e434ed8bfb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c198fd2a68ae4621f49da32241b8608f25500c64fda998ff30a2827992e7896
+size 81911
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..c668c204a94d10997108d0ac0a9f07765d00037e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video presents a stunning view of Earth from space, showcasing the continents and oceans in a realistic and detailed manner. The image is a composite of three frames, each capturing a different perspective of our planet. The first frame provides a broad view of the Earth, highlighting the vastness of our planet and the intricate patterns of the oceans and landmasses. The second frame zooms in on the Americas, revealing the diverse landscapes and the intricate network of rivers and lakes. The third frame offers a closer look at the United States, showcasing the country's unique geography and the intricate patterns of the states. The video is a testament to the beauty and complexity of our planet, captured in a realistic and detailed manner."
+    ],
+    "video_ids": [
+        "4TgB5tDbElw_23_0to130"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Continents, Oceans, Americas, United States? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the continents of North and South America, including the United States as part of North America. The oceans surrounding these continents are also visible, fulfilling the described 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a Earth and the and South America, which the United States, part of the America. The oceans surrounding these continents are also depicted, fulfilling the ' 'O(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e9eabc51aa505e57f271abf65bd5ddb25175e249
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dce05d38ca50eb368c3baa987b4f3b2a02048abd46f012d03101bdc8181cc26d
+size 297158
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..dc69e133a2f0661712ffeca2525078ee70373e0a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a cooking tutorial for making Bacon and Egg Fried Rice. The style of the video is a close-up, step-by-step demonstration of the cooking process. The main focus is on a metal baking dish filled with cooked rice, which is being prepared for the next step in the recipe. The dish is placed on a wooden table, and there are other ingredients visible in the background, including bacon and eggs. The video is likely to be informative and engaging, with clear instructions and visuals to guide viewers through the cooking process."
+    ],
+    "video_ids": [
+        "LVnxakIqKf4_2_0to123"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Metal baking dish, cooked rice, wooden table, bacon, eggs? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a metal baking dish filled with cooked rice, placed on a wooden table. There is also a plate of bacon visible, and although eggs are not directly visible, the text overlay mentions 'Bacon and Egg Fried Rice,' implying eggs are part of the recipe. The core objects described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a metal baking dish containing with cooked rice mixed which on a wooden table. There are bacon bacon visible with bacon and in which the eggs are not directly shown, the context ' suggests 'egacon, Eggs', Rice', suggesting the are part of the dish. The presence elements ( in present, contradiction.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..650e3b29dc2126238976be8c104d3a7474ee6e82
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3307ff21c53eea31a4b04d0f1dc2b77a4a7c9c207582c5ecef8e396c6d06520c
+size 129599
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..b305b9c99d77c431d7f0268f4bcebbb321949ba9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a dessert being eaten. The dessert is a slice of cheesecake, which is being scooped up with a spoon. The cheesecake is placed on a plate, and there are pieces of fruit, including strawberries and kiwi, arranged around it. The plate is on a table, and there is a bowl in the background. The style of the video is a simple, straightforward food video, focusing on the dessert and the process of eating it. The lighting is bright, and the colors are vibrant, highlighting the textures and colors of the food. The video does not contain any text or additional elements, and the focus is solely on the dessert and the act of eating it."
+    ],
+    "video_ids": [
+        "Ml5AAM_O6lE_46_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Cheesecake slice, spoon, strawberries, kiwi, plate, bowl? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a cheesecake slice, a spoon scooping from it, strawberries, kiwi, a plate, and a bowl in the background. All specified objects are present and identifiable in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showing showing a cheesecake slice, a spoon,oping the it, strawberries, kiwi slices and plate holding and a bowl in the background. The the objects are present and identifiable in the scene.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3e475b1f220ef2247d877ea2b4b550e1af448119
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1318-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dbe9f847c6d9318550a4c8429372f934b3a74817c6d461a26a3a301cd0ecbf5a
+size 88553
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..568533b8ded4547acd724e9a0e7c1c6ef947a51d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of triumph for a soccer player. The player, dressed in a blue jersey, is seen celebrating a goal with a fist pump. The background is a blur of a soccer stadium, indicating the setting of a professional match. The player's joyous expression and the intensity of the celebration suggest a significant victory. The video is a dynamic snapshot of the player's achievement, encapsulating the excitement and passion inherent in the sport of soccer."
+    ],
+    "video_ids": [
+        "tTm1swUEYSk_38_0to145"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A soccer player in a blue jersey performing a fist pump.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a blue jersey, and he is clearly performing a fist pump gesture. The player's attire, the action, and the context (implied by the blurred stadium background) align with the description. Additional elements like tattoos or sponsor logos do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person player in a blue jersey, which the is performing performing a fist pump.. The player's attire and the action, and the setting ofaplied by the setting background background) all with the description of There elements like the or other logos on not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..87e4bc218006eee22989489a7b16870ad53c599b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d12486878df4c0a9419e458b2df795b1d7a4b3487967884f39b41445102db512
+size 231975
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..1911b44a01d1400ac75d6842a5eb93ece98277bf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a peaceful walkway lined with lush green trees and bushes, leading towards a row of apartment buildings. The walkway is paved with brick, and the sun casts a warm glow on the scene, creating a serene atmosphere. The apartment buildings are modern and well-maintained, with balconies and windows that reflect the sunlight. The overall style of the video is calm and inviting, capturing the beauty of urban greenery and residential architecture."
+    ],
+    "video_ids": [
+        "MsxJb0Y_H0Q_39_101to238"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Lush green trees and bushes, brick-paved walkway, apartment buildings with balconies and windows.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing lush green trees and bushes lining both sides of a brick-paved walkway, with apartment buildings featuring balconies and windows visible on the left side. The scene matches the described elements without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing lush green trees and bushes, a sides of a brick-paved walkway. with apartment buildings featuring balconies and windows in in the right side of The presence is the description elements without any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fe7b3b3fbae12d44684e64c5e14ba06d884e56bc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:718440423e2dfb80f1d0eb21e7aacdf261e4f0afd7cc4e394b2a4b3a884b61d4
+size 219940
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..bafea9ab06ef70538294de660ce3179c3c871c60
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man enjoying a meal at a restaurant. He is wearing glasses and a striped shirt. In the first frame, he is holding a spoonful of food to his mouth. In the second frame, he is taking a bite of the food. In the third frame, he is looking at the camera with a smile on his face. The food on his plate appears to be a bowl of rice with some vegetables. The restaurant has a floral wallpaper on the wall behind him. The overall style of the video is casual and relaxed, capturing a simple yet enjoyable moment of dining."
+    ],
+    "video_ids": [
+        "BozPRwYOJrw_28_0to145"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a spoon, a bowl of rice, some vegetables, the man's glasses, and a striped shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing glasses and a striped shirt, eating from a bowl of rice with a spoon. The bowl contains rice and some visible vegetables (like green onions and possibly shrimp). All specified objects are present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing glasses and a striped shirt, eating from a bowl with rice with a spoon. There bowl contains rice and some vegetables vegetables.possibly green beans). possibly a). The the elements are present and correctly depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2e3154451ec7143340825252339d0be157c0c6b2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:10b4b3905e40b0d91d85bd28f045337aec0243a1dc72913e0254eb1e7a21634c
+size 308794
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..ce260004d365349360801756850c7d992be0658c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a car's side vent, featuring a logo with an orange and black design. The logo is a stylized letter 'S' with a shield-like shape around it. The vent has a black background with a series of horizontal slits. The car's body is a light gray color, and the vent is located on the side of the vehicle. The style of the video is a straightforward, static shot with no movement or action. The focus is on the design and details of the car's vent and logo."
+    ],
+    "video_ids": [
+        "_fMt7CG0b40_3_0to138"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Car's side vent, logo (stylized 'S' with a shield-like shape).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a car's side vent with a stylized logo resembling a shield shape, which matches the description. The logo is prominently displayed within the vent, and there are no conflicting elements that contradict the specified object.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a car's side vent with a logoized ' that a shield with containing featuring includes the description of The logo features prominently displayed in the vent, and the are no additional elements in contradict the given object(s}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a29a028e8150f60e0d993f28e91466bb0ca0c4ae
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75cbf6dc43825a518ba1cdf6daaaca7101e6b6ec9d4c17440415ce2dec505c07
+size 90372
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..ac31a912e709642bca1020a01ab9fbaa70000bc5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a large red snowmobile driving across a snow-covered landscape. The snowmobile is equipped with a large plow attachment at the front, which it uses to clear a path through the snow. The driver is visible through the windshield, wearing a helmet and goggles for protection against the cold and snow. The snowmobile is moving at a steady pace, leaving a trail of disturbed snow behind it. The landscape is flat and featureless, with no other objects or landmarks visible. The sky is clear and blue, suggesting a cold but sunny day. The style of the video is realistic and documentary, capturing the snowmobile in action without any artistic embellishments."
+    ],
+    "video_ids": [
+        "gJFwdYnyjyg_6_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large red snowmobile with a front plow attachment and a driver inside.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large red vehicle with a front plow attachment, which appears to be a specialized snowmobile or snow tractor, and a driver is visible inside the cab. The vehicle is operating on a snowy, icy surface, consistent with the described scenario. Additional elements like the black circular object and steam from the vehicle do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large red vehicle with a front plow attachment, which matches to be a snow snowmobile or snow groom. moving there driver inside visible inside the vehicle. The vehicle is travers on a snowy landscape open terrain, which with the ' '. The elements like the clear and object on the-like the exhaust's not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9b5faa6c9bd7ee9ad549f5305fba38bb372ef275
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60316d1d107241054f1bb44c1c6ae44b0f6932e8cf9c55fb23007641676f0930
+size 253388
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..1862fe9bd2f04d65442859e8291efebf40b1722c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen in a vintage setting, sitting at a desk with a typewriter. He is wearing a dark blue shirt and glasses, and appears to be in the process of typing. The desk is situated in front of a yellow curtain, and there is a chair visible to the left of the frame. The man's focus is on the typewriter, and he seems to be deeply engrossed in his work. The overall style of the video is reminiscent of a bygone era, with the typewriter and the yellow curtain adding to the vintage feel. The man's actions suggest that he is a writer or a journalist, working diligently on his latest piece. The video captures a moment of quiet concentration and creativity."
+    ],
+    "video_ids": [
+        "GBbsNKaVAB0_10_0to130"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a desk with a typewriter, a yellow curtain, and a chair.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man interacting with a typewriter placed on a small table (which can be considered a desk), with a yellow curtain in the background and a chair visible to the left. These core elements are all present and consistent with the description, even though the man is not seated in the chair and the typewriter is on a round table rather than a traditional desk. No elements contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man sitting with a typewriter on on a desk desk.which can be considered a desk). with a yellow curtain in the background. a chair visible to the left. The elements elements match present present and match with the description. fulfilling though the typ is not sitting on the chair.\"\n the tableewriter is not a table table rather than a standard desk. The additional contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..14186643bd8f1dda3c48148b0bffa05ef4b3d0f2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:287f8c6ae2ce0a5f91ccbcbaefe04c88e21850100318e6b7f09b3bac665e196c
+size 85535
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..f96bdc279572c7de60c210e8f3799db229926608
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a polar bear in an enclosure at what appears to be a zoo or wildlife park. The bear is standing on its hind legs, leaning against the glass barrier of its pool, which is filled with clear blue water. The bear's fur is thick and white, contrasting sharply with the blue water and the grayish rocks at the bottom of the pool. The bear's front paws rest on the glass, and it seems to be looking out towards the visitors or perhaps the other side of the enclosure.  As the video progresses, the bear shifts slightly, maintaining its position against the glass. The background reveals a covered area with a white canopy, suggesting a shaded space for the bear. Beyond the canopy, there are trees and parts of the zoo's infrastructure visible, indicating a well-maintained environment designed to mimic the polar bear's natural habitat.  The camera remains stationary throughout the sequence, focusing on the bear and its immediate surroundings. The water in the pool is calm, with gentle r"
+    ],
+    "video_ids": [
+        "f4be3c0e4cdf311b0da4a996c36e30eb9711dbb70dfeb3186d49d64ed88b0dd6"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A polar bear? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a polar bear interacting with a glass barrier in an aquatic enclosure, which matches the description of the object. The bear's white fur, size, and behavior are consistent with a polar bear, and there are no elements contradicting this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a polar bear, with its pool barrier in a environment environment. which align the description of the ' being The bear's white fur, posture, and posture are consistent with the polar bear. and the are no elements ining this description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..59f5a14fd56faae60b7df07a4d53baeb811ddf26
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:74458a0988829467df70724687d2671a5d6c1bcd0b3c7a34c217687efc8e82b1
+size 209641
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..28fe049775e9e8c91dfa568267fdececa4bbe156
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a nature documentary that follows a man in a straw hat and blue shirt as he walks through a lush, green garden filled with vibrant flowers and trees. The man is carrying a tray of plants, suggesting that he might be a gardener or a farmer. The garden is located on a hillside, providing a beautiful view of the ocean in the background. The man's journey takes him through a variety of plant life, including a large bush with purple flowers. The overall style of the video is serene and peaceful, capturing the beauty of nature and the man's connection to it."
+    ],
+    "video_ids": [
+        "Cj4t6p4M-hY_10_16to139"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a straw hat and blue shirt, a tray of plants, a large bush with purple flowers.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a straw hat and blue shirt carrying a tray of plants. There is also a large bush with purple flowers visible in the foreground. These elements match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man wearing a straw hat and a shirt, a tray of plants. He is a a large bush with purple flowers in in the background. The elements match the description provided, the questionObject(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5aacf5cd848d8d0a5005d09d1788e9096e8d9da3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1319-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a07d32e0471328a081ba0f32acc67d651c9cc456aa11936143b1e8051f260d6
+size 285978
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..46a717db0950c5eaed56fca3cf7abfe7ff91d8d5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with glasses, wearing a black top, standing in front of a television screen displaying the logo of a news channel. The woman appears to be speaking or about to speak, as she is slightly open-mouthed. The television screen is mounted on a wall with a blue and orange color scheme. The room has a modern and professional look, with shelves in the background containing various items, including books and what appears to be a model of a giraffe. The overall style of the video suggests it could be a news segment or a television show."
+    ],
+    "video_ids": [
+        "4846hAPP9iQ_0_0to150"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with glasses, wearing a black top, and a television screen displaying a news channel's logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with glasses wearing a black top, standing in front of a television screen displaying the 'D NEWS' logo. The core elements described are present and accurately represented. Additional background elements (like shelves with scientific items) do not contradict the description and are acceptable as they do not conflict with the core object condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing glasses wearing a black top, and in front of a television screen. a logoCNNuck' logo, The description elements of in present, accurately represented in The elements elements likelike the with decorative models) do not contradict the main.\"\n are acceptable.\"\n they do not interfere with the main elements condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2b80d84ff8a47b0eb2ed36be53c75ae31532c2f2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:14a839017004d5a3713b6552b587f7bbb57ed8c036d08b3c404fb2b1af9cf38f
+size 80776
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..7548bf356bdd8f4d6e982463afbe0d48175a32dd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a basketball player in action, showcasing his athletic prowess and team spirit. The player, dressed in a vibrant purple Lakers jersey, is seen flexing his muscles in a triumphant pose, his arms raised high in the air. The background reveals a bustling basketball court, complete with a hoop and a scoreboard, indicating an ongoing game. The player's intense focus and the dynamic environment suggest a high-stakes match, filled with excitement and competition. The video encapsulates the essence of the sport, highlighting the player's dedication and the thrilling atmosphere of a basketball game."
+    ],
+    "video_ids": [
+        "uMTmOVTcJBU_13_148to289"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A basketball player in a vibrant purple Lakers jersey, flexing his muscles in a triumphant pose with arms raised high in the air.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a basketball player wearing a vibrant purple Lakers jersey with the number 12, flexing his muscles and raising his arms high in a triumphant pose. The core description is accurately fulfilled, and additional background elements (like the basketball hoop and crowd) do not contradict the main subject's depiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a basketball player wearing a vibrant purple Lakers jersey, the number 2 on. flexing his muscles in raising his arms in in a triumphant pose. The setting elements of largely represented, with the elements elements likelike the basketball court and court) do not contradict the main subject.\"\n actions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..41cc3706ad21af74644fe428c120c84a9d4a133a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f645fb08b4aaf890dd3e7d009779206902d87e5b6403fe816fd95b1f06f728d
+size 238311
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..4511dddde8533fc631d055ac3777c56acd7b5a69
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men standing in a large, open room with a green wall and a large window in the background. The man on the left is wearing a blue jacket with yellow stripes and a grey shirt underneath. The man on the right is wearing an orange hoodie with a black jacket over it. Both men are looking at each other and appear to be engaged in a conversation. The room they are in has a high ceiling and is quite spacious. The lighting in the room is bright and natural, coming from the large window. The overall style of the video is casual and candid, capturing a moment between the two men."
+    ],
+    "video_ids": [
+        "aeqHxlnfThw_48_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men - one on the left wearing a blue jacket with yellow stripes and a grey shirt; one on the right wearing an orange hoodie with a black jacket.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men standing in what appears to be an industrial setting. The man on the left is wearing a blue jacket with yellow stripes and a grey shirt, matching the description. The man on the right is wearing an orange hoodie under a black jacket, also matching the description. Their attire and positions are consistent with the specified 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men standing face a appears to be an indoor or. The man on the left is wearing a blue jacket with yellow stripes and a grey shirt, while the description. The man on the right is wearing an orange hoodie with a black jacket, also matching the description. The attire and the align consistent with the given 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c6313e82771700794d2e1efea01b515b8ce3c641
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6aa553a1636a9d6badce09387f927209c50e775bb82c7d9559dde875d739a357
+size 163573
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..70fbae310d0f532984b6a2df52c515e334f1bbb2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a cooking tutorial featuring a woman in a kitchen. She is wearing a red top and a necklace. The kitchen is well-equipped with wooden cabinets, a sink, a stove, and various cooking utensils. The woman is smiling and appears to be in the middle of a cooking process. The video is likely to be informative and engaging, with a focus on the cooking process and the woman's interaction with the kitchen environment."
+    ],
+    "video_ids": [
+        "cIJ0k-YdfjA_19_0to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman wearing a red top and a necklace, in the middle of a cooking process, smiling.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The woman is wearing a red top and a necklace, and she is smiling while standing in a kitchen. Although she is not actively engaged in a cooking process in the visible frames, she is positioned in a kitchen setting, which aligns with the general context of a cooking process. The core description is largely fulfilled.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video in wearing a red top and a necklace, and she is in. standing in a kitchen. There the is not actively cooking in cooking cooking process, the video frames, the is positioned in front kitchen setting with which aligns with the description context of cooking cooking demonstration. The presence elements is largely matched by}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7ababb6c77ce28a9a50132cafcaeec65df1460e0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48b371ad3e339eea0d487f29bc7d7d52cef7367f4b2e329e3de4a451314ae01d
+size 80930
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..0cf686866fc844815ca90ccf3766241df11b66b6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a cartoon character with a red and purple hairstyle is standing in front of a television screen. The screen displays a vibrant image of a volcano. The character is holding a pink donut and is pointing at it with a smile on their face. Another cartoon character, wearing a green hat, is standing next to the first character, looking at the donut with a curious expression. The background features a colorful, whimsical landscape with a purple sky and a large, purple tree. The overall style of the video is playful and imaginative, with a focus on the interaction between the characters and the vibrant visuals."
+    ],
+    "video_ids": [
+        "_2tDzgB03lk_56_71to242"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A cartoon character with a red and purple hairstyle, a television screen displaying a vibrant image of a volcano, a pink donut, another cartoon character wearing a green hat, and a large, purple tree.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a cartoon character with a red and purple hairstyle (the boy), a television screen displaying a vibrant image of a volcano, a pink donut (which is being tossed and caught), another cartoon character wearing a green hat (the girl), and a large, purple tree (with yellow flowers) in the background. All specified objects are present and accurately described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video contains fulfills the 'Object(s)' condition as depicting depicting a cartoon character with a red and purple hairstyle,the character on a television screen displaying a vibrant image of a volcano, a pink donut beingheld the being held by caught by another cartoon character wearing a green hat (the girl), and a large, purple tree inin a leaves) in the background. The elements elements are present and contribute depicted.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..941e9f121524517ef58b49490bf3b203546895db
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4043d19113b21213c440f35c3f9e54553809c51459c442029ba321c363dee12
+size 140841
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..4cab6fcd527bff7f027c9fef6052a36cedbd970f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a motorcycle engine, focusing on the shiny chrome parts and the intricate details of the engine components. The engine is open, revealing the internal parts, and the motorcycle is parked on a concrete surface. The style of the video is a time-lapse or a series of still images, capturing the motorcycle from different angles and perspectives. The lighting is bright, highlighting the metallic surfaces and the reflections on the chrome parts. The video does not contain any text or additional elements, focusing solely on the motorcycle engine."
+    ],
+    "video_ids": [
+        "2HuYuTXwalw_9_0to194"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Motorcycle engine, chrome parts, internal engine components? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a motorcycle engine with visible chrome parts and internal engine components, such as pipes, bolts, and mechanical assemblies. The focus remains on these elements throughout the frames, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a motorcycle engine, chrome chrome parts and internal engine components. such as the and fins, and the parts. The presence is on the elements, the frames, fulfilling the 'Object(s)' condition as}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e8bf91977905802d3c55929ae21c01ed66e013f1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37c6ab38b9ba58d5d48f17d75e5c6f5e212ba04fbb5ce76848c9c00f95adaf36
+size 205042
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..eba9f0b00f53dda9c842cc80a55122222229e771
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated sequence featuring a cartoon baby character. The baby is depicted in a playful and whimsical setting, with a vibrant pink background adorned with stars and clouds. The baby is shown in three different positions: sitting, standing, and lying down. The baby is wearing a white diaper and has black hair. The baby is surrounded by colorful elements such as a rainbow, a candy cane, and a flower. The overall style of the video is cute and cheerful, with a focus on the baby's movements and expressions."
+    ],
+    "video_ids": [
+        "2oopbQOW7WM_55_0to117"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Cartoon baby, rainbow, candy cane, flower? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly features a cartoon baby crawling on a colorful surface, with a rainbow visible in the background, candy canes on the left side, and flowers on the ground. These elements are consistent with the 'Object(s)' condition described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts cartoon cartoon baby, on a cloud background, which a rainbow and in the background. a canees, the left side, and flowers on the right. The elements align all with the 'Object(s)' condition provided.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..68b3e9f250d32c1f9d87d41b27317c61f5f42e0e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3eddf6615f3e6efa1ed413522534d6ad8343c2af2792de896facc5816de8db4
+size 187653
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..6d1df7c72f1f112194d7ce13f8c1845402a3517d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a gray tabby cat lying on a red blanket. The cat appears to be relaxed and comfortable, with its head resting on the blanket. The cat's fur is striped, and its eyes are open, looking directly at the camera. The background is blurred, but it seems to be an indoor setting with a chair and a table. The style of the video is a close-up shot of the cat, focusing on its facial expression and the texture of its fur. The lighting is soft, creating a warm and cozy atmosphere. The video captures the cat's peaceful demeanor and the comfort of its resting place."
+    ],
+    "video_ids": [
+        "GqsQ-c7byhU_2_0to113"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Gray tabby cat, red blanket? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a gray tabby cat resting on a red blanket, which matches the specified 'Object(s)' condition. The cat's fur pattern and color, along with the blanket's color and texture, are accurately depicted without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a gray tabby cat with on a red blanket. which matches the description objectsObject(s)' condition. The cat's fur pattern and the, as with the red's color, texture, are consistent depicted, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a83868f04ea3ebaca28268f46981cefd54bc2bd4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-132-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e16fa0e7dffa4962698cddb556e70f0b357b431497147498e71e4e50543378d4
+size 215476
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..a14af2fa3b0cfb25f0a3c6d92a480870bf6aa325
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a large, menacing gorilla holding a stick and walking towards the camera. The gorilla's fur is dark and shaggy, and its face is contorted in a fierce expression. The gorilla's eyes are wide and intense, and its mouth is open as if it's growling or roaring. The background of the video shows a jungle-like setting with trees and foliage. The lighting in the video is dim, creating a sense of mystery and danger. The gorilla's movements are slow and deliberate, adding to the overall sense of tension and suspense. The video is likely a scene from a movie or TV show, and the gorilla is likely a character or a creature that the viewer is meant to be afraid of."
+    ],
+    "video_ids": [
+        "dbjl4Wa1FIE_3_390to523"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large, menacing gorilla holding a stick.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large, menacing gorilla holding a stick, which matches the core description. The gorilla's aggressive expression and the stick it is holding are clearly visible. While there are additional elements like the background structures and what appears to be the head of another animal in the foreground, they do not contradict the primary object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a large gor menacing gorilla with a stick, which align the description description. The gorilla's posture posture and the stick in is holding are consistent visible, The the may no elements like the forest foliage and the appears to be a edge of another gor, the background, these do not contradict the main focus description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bb70ca1109cb9daa21b41e1f335565ad4c2deb84
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:011a2600438bd09db2109b82aad14a794a8a93cb00edde8e4d4ec86aa3bfff03
+size 297442
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..0f3c70c0828e5a9df078e0133d1a60e40dc08da0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a textured surface, likely a wet pavement or a similar material, with intricate patterns created by water droplets and streaks. The surface is predominantly blue with white highlights, suggesting it might be a painted or treated material. A small insect, possibly a bee, is seen crawling along a thin, dark line that cuts across the surface. The bee's movement is slow and deliberate, indicating it may be navigating through the water or searching for something specific. The camera remains stationary throughout the sequence, focusing on the bee's journey across the textured background. The lighting is natural, casting subtle shadows that enhance the three-dimensional quality of the scene. There are no other significant objects or characters present in the frame, keeping the viewer's attention solely on the bee and its interaction with the environment."
+    ],
+    "video_ids": [
+        "748324cc5df99675165e68ebd5af9d6b80b4404bc4b32678b5f94632a8054a2c"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small insect (possibly a bee) crawling along a thin, dark line.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small, fuzzy, brownish insect that resembles a bee, positioned on a thin, dark line that appears to be a crack or seam in a textured surface. The insect is actively crawling along this line, matching the core description. The surrounding surface has a blueish, textured appearance, but this does not contradict the main subject's behavior or identity.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a close, dark object yellow object object-like appears a bee, crawling along a thin, dark line within appears to be a vein or a in a textured,. The surface is moving moving along this line, which the description description of The rest environment is a metallic and, metallic appearance with which it does not contradict the presence focus of presence and presence.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0ab2d8ffbc19b64ba06814a9e4933161e561b197
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:455e60e995f4f7ffff719727067d9e95e62efee1ce9abce0f0e6224e4539f2f9
+size 247957
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..bffd94354712ede157097eec293308e45948a090
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is on two players from the same team, both dressed in white and blue uniforms with the number 3 prominently displayed on the back of one player's jersey. The player with the number 3 is in the process of throwing a pass, his arm extended and the football visible in his hand. The other player, wearing a helmet with the number 6, is on the ground, possibly having just made a tackle or fallen during the play. The background is filled with other players, some in motion and others standing still, all dressed in similar uniforms, indicating that they are part of the same team. The style of the video is a real-time capture of the action, with no visible filters or artistic effects, providing a clear and unobstructed view of the game in progress."
+    ],
+    "video_ids": [
+        "qvjjYjIf1b8_17_0to155"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two players (one with number 3, the other with number 6), other players? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two players: one kneeling with number 6 on his jersey and another standing with number 3. The standing player is clearly visible, and although other players are blurred in the background, they do not contradict the core description. The focus remains on the two specified players, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a players, one wearing with the 3 and their jersey and another standing with number 3 on There presence player is in visible, and the the players are present in the background, they do not contradict the core description. The presence is on the two players players, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a385ba9b17ac8f14e15acc221f4a5e7af031f866
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1299bb91c8c12f257f1cb1a67d8d3dd8b94bbdb6358a4a4f253c0f3e39afab7
+size 283220
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..c3f59c5a9be86a1bc63529c04271efc454b47736
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young woman with long brown hair is seen in a kitchen, preparing a meal. She is wearing a white shirt and is standing in front of a wooden counter. On the counter, there are several bowls and bottles, as well as a blender. The woman is holding a spoon and appears to be stirring something in one of the bowls. The kitchen is well-lit, with natural light coming in from a window. The overall style of the video is casual and homey, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "9jKtf27L2u8_9_90to254"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman, several bowls, bottles, a blender, a spoon? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a young woman standing in a kitchen setting. Several bowls (metal and white) are visible on the counter in front of her. Bottles are present in a rack on the counter behind her. A blender is also visible on the counter. Although a spoon is not clearly visible in the provided frames, the presence of other specified objects confirms the video largely matches the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a young woman in in a kitchen,. She bowls arewhitelic ceramic) are visible on the counter, front of her. Thereles are also in the cabinet behind the counter behind her. A blender is also visible on the counter. The the spoon is not explicitly visible in the woman frames, the woman of a kitchen objects like that video largely matches the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..199f4b8418d2723423721ff79ce24466560d65dd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b9f3e874336153ca053e6ce1ad88e570e482ee3108dfa3436a77dcf6c0c4f3bd
+size 159299
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..837275ed9500530a7afa7dfce167f9ebbb7ff4a6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a humorous scene of a group of people attempting to move a large, oddly shaped object, which appears to be a giant bird wing, across a parking lot. The wing is so large that it nearly covers the entire frame of the image. The group consists of three individuals, each dressed in casual attire and wearing winter hats, suggesting a cold weather setting. They are positioned in various stages of movement, with one person leading the way, another following closely behind, and the third person trailing behind. The wing is being carried over a blue sports car, which is parked in the background. The car is sleek and shiny, with a distinctive design that stands out against the otherwise mundane surroundings. The parking lot is expansive, with a few other cars parked in the distance. The overall style of the video is candid and unscripted, capturing a moment of everyday life with a touch of whimsy."
+    ],
+    "video_ids": [
+        "FPQ4zgkWgIU_45_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large, oddly shaped object (giant bird wing), three individuals, a blue sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a large, oddly shaped object resembling a giant bird wing, three individuals (two standing and one crouching), and a blue sports car. All core elements from the description are present without contradiction, even though the scene is clearly AI-generated and surreal.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a large, oddly shaped object ( a giant bird wing, three individuals interactingtwo people and one walkingrouching), and a blue sports car. The elements elements of the description are present, any.\"\n and though the context appears AI AI-generated and the.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3bbd98e850162b7208e3b4cfa93c0ce2f3295819
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b54bd07f90af62d63c6c7b222e29ccc9ac9980bb4617b158834cac3da3a994ec
+size 197233
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..8797556994aa9fb78f43db2e3ee010d5f83e5375
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the iconic Eiffel Tower in Paris, France, from a low angle perspective. The tower, constructed from an intricate lattice of metal, stands tall against a backdrop of a clear blue sky with a few scattered clouds. The sun casts a warm glow on the tower, highlighting its intricate details and the play of light and shadow. The video is a time-lapse, showing the tower in different lighting conditions as the sun moves across the sky. The overall style of the video is a blend of architectural beauty and natural light, creating a visually stunning representation of this world-renowned landmark."
+    ],
+    "video_ids": [
+        "1qsZbzDftFA_1_0to159"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Eiffel Tower? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the Eiffel Tower, capturing its iconic iron lattice structure and towering presence against a partly cloudy sky. The perspective, focusing on the base and ascending structure, accurately represents the landmark as described in the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently depicts the Eiffel Tower, which its iconic structure lattice structure against the presence against a blue cloudy sky. The image and lighting on the tower and lower towards, is represents the E. described. the question.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b2dee9d0b6ab3522b0049a82cf9e1e0b47e36fb7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be667ec4c5fa343241f29288c1923d5d737d4b83fde36f1aec11d46c483b1437
+size 286283
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..9cfff48c94db6ced7d47cd1a27bf58cb9f9d43ff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a garage, working on a car. He is wearing a black t-shirt and blue gloves. The car is lifted up on a lift, and the man is looking at the engine. The garage is well-lit, and there are tools and equipment visible in the background. The man appears to be focused on his work, and the video captures the process of car maintenance."
+    ],
+    "video_ids": [
+        "PuFxUmpd7Ks_30_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a car lifted on a lift, tools and equipment.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man standing under a car that is lifted on a lift, which matches the core description. Although tools and equipment are not clearly visible, the setting (a garage) and the man's attire (gloves) imply a mechanical or repair context where tools are likely present. The core elements are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man working next a car that is lifted on a car. which ful the ' description of The the and equipment are not explicitly visible, the setting anda garage or and the man's actions (suggestloves) suggest that maintenance or automotive context, such and likely present. The presence elements of present represented.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..09476bb0be6e91d2f049a131c8fec553031ceaab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:409b642a1111328ecd9d1ab8fff4cf9aee0a38edaee05ac670a42994d670eddb
+size 166789
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..0f578220c281596a3ca4c65d524506cfb03a58d3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up of a plate of food, showcasing a delicious meal. The meal consists of a Denver omelette, which is a type of omelette filled with ham, onions, and green peppers. The omelette is golden brown and appears to be cooked to perfection. Accompanying the omelette is a side of hash browns, which are crispy and golden brown. There are also two slices of toast on the plate, which are lightly toasted and appear to be made from whole wheat bread. On the side, there is a fresh orange slice, adding a pop of color and a hint of freshness to the meal. The overall style of the video is simple and straightforward, focusing on the food and allowing the viewer to appreciate the details of the meal. The colors are vibrant and the textures are varied, making the meal look even more appetizing. The video does not contain any text or additional elements, keeping the focus solely on the food."
+    ],
+    "video_ids": [
+        "9CM_uzHNvbQ_12_0to120"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Denver omelette, hash browns, toast, fresh orange slice? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully displays all the specified items: a Denver omelette (with visible ham and green peppers), hash browns, toast, and a fresh orange slice. These elements are clearly identifiable and presented together on a plate, matching the description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a the mentioned objects: a Denver omelette,with visible ham, cheese peppers), hash browns, toast, and a fresh orange slice. The elements are clearly visible and match on on a plate, matching the description provided any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..87f368c429295381ee8f120a4b86dea2b838a296
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1320-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8bbd0636528e247279ecaa229213d12c4d85091ab150c87ece931a6f1b4a15fa
+size 121542
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..54bc13af6d827d3a69daba33646e02c6851a9a89
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a small brown dog is seen standing on the hood of a blue car. The car is parked in a lot with other cars visible in the background. The dog appears to be curious about the car and is looking around. The car's hood is open, revealing the engine underneath. The dog seems to be exploring the car, possibly attracted by the warmth of the engine. The scene is set in a parking lot, and the dog is the main focus of the video. The dog's actions suggest a sense of adventure and curiosity."
+    ],
+    "video_ids": [
+        "7CdhV-Drtgc_8_0to104"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small brown dog? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small brown canine-like animal standing on a car hood, which matches the description of a 'small brown dog'. While the animal's exact species is not confirmed, its appearance and behavior are consistent with that of a dog, and there are no elements contradicting the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small brown dog, figure inside inside the car's, which align the description of a 'small brown dog'. The the animal's exact breed is not definit, it size and size align consistent with a of a dog. fulfilling the are no elements ining this description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..add479494c468af8cc3fb589177853fd180973d1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ee834b3b0883408b53180c3669bea848a4f5df87f7ae0f7047fb033dac2a924
+size 240697
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..9dfd41a2f04d28cfe0421dd6d483075701dbf200
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a close-up view of a small, clear aquarium filled with water. Inside the tank, there is a variety of colorful decorations, including artificial plants, rocks, and a small red structure resembling a bridge or a house. Several small fish, displaying vibrant colors such as yellow and orange, swim gracefully around the tank. The water is clear, allowing for a detailed view of the decorations and fish. A transparent tube is inserted into the water, from which bubbles rise to the surface, creating a dynamic visual effect. The background features a green bottle with an orange label, partially visible behind the aquarium. The camera remains stationary throughout the video, focusing on the aquarium and its contents, capturing the serene movement of the fish and the gentle rise of bubbles. The text \"Craft Master\" appears in the bottom left corner of the frames, indicating the creator or brand associated with the video."
+    ],
+    "video_ids": [
+        "31be31e3b57a87ceadf9e4ca1f2bb6fe3b9a7e15b44cd595a9debe32c495e96d"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Artificial plants, rocks, a small red structure, several small fish, a transparent tube, and a green bottle with an orange label.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows artificial plants, rocks, a small red structure (which appears to be a bridge or ladder), several small fish swimming, a transparent tube (likely part of a water circulation system), and a green bottle with an orange label in the background. All elements described in the condition are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts an plants, rocks, a small red structure,which appears to be a house or house), several small fish,, a transparent tube (which part of the water filter system), and a green bottle with an orange label in the background. The these match in the condition are present in match depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5ceade637b831ed81bd840997d822f1e1577550e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0e89401cdd5a944b2aea9818762980f9dc8c4eb34a87c3ed77897e9c036aac34
+size 189755
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..8da1c7c9601ca4b134194e8fa2d4c85efc3f2d64
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a car's side vent, featuring a logo with an orange and black design. The logo is a stylized letter 'S' with a shield-like shape around it. The vent has a black background with a series of horizontal slits. The car's body is a light gray color, and the vent is located on the side of the vehicle. The style of the video is a straightforward, static shot with no movement or action. The focus is on the design and details of the car's vent and logo."
+    ],
+    "video_ids": [
+        "_fMt7CG0b40_3_0to138"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Car's side vent, logo (stylized 'S' with a shield-like shape).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a car's side vent with a stylized logo resembling a shield shape, which matches the description. The logo is prominently displayed within the vent, and there are no conflicting elements that contradict the specified object.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a close's side vent, a stylized ' that a ' with with which is the description. The vent is not displayed in the vent, fulfilling the are no additional elements in contradict the given objects(s}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7807be4868b9060e832a32f3801964065dcbd27e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64c6ee812bc3428e55852f96784955d6b10bf92d60e9038132f56ee672da95f5
+size 50084
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..3c96dba47a0b0bb5d50094c815f1f9747a2f4f45
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a small, dark-colored animal, likely a bat, lying on a textured gray surface, possibly concrete or asphalt. The bat is positioned centrally in the frame, with its body slightly curled up. Its fur appears wet, suggesting it may have been exposed to water or rain. A white object, resembling a piece of paper or a small hat, rests on top of the bat's back. The bat remains still throughout the sequence, and there are no significant changes in its position or the surrounding environment. The lighting is dim, creating a somber atmosphere. The camera maintains a steady, fixed position, focusing solely on the bat and the white object without any noticeable movement or shift in perspective."
+    ],
+    "video_ids": [
+        "c53b8ced338590a60eda992c3318ae375a9929d74ef2766de2e8db290ef857ac"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small, dark-colored bat and a white object resembling a piece of paper or a small hat.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small, dark-colored object that resembles a bat, lying on a gray surface. On top of it is a white object that looks like a piece of paper or a small hat. The description is largely accurate, and no conflicting elements are present that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a dark, dark-colored bat that resembles a bat, and on a surface surface. There top of this, a white object that looks like a piece of paper or a small hat. The description matches largely matched, and the additional elements are present.\"\n would the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7f4472fb8193570419df001752e617e19a13ffb3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b33ba530cca204a3a1ffe3df10acc5ae0886399c4ef76bed0e778ec78b7c00e
+size 178327
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..a93423f422d37d5769c57e882953ec511a96e480
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a couple's dance in front of the Cloud Gate sculpture, also known as \"The Bean,\" located in Millennium Park, Chicago. The couple, dressed in formal attire, is seen dancing in a graceful and elegant manner. The sculpture's reflective surface mirrors the city skyline and the couple's dance, creating a captivating visual effect. The video is shot in a way that emphasizes the beauty of the sculpture and the joy of the couple's dance, capturing a moment of celebration and love in an iconic urban setting."
+    ],
+    "video_ids": [
+        "WPGyu-fTWYU_28_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A couple dressed in formal attire? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a couple dressed in formal attire\u2014a woman in a white wedding dress and a man in a white suit with a bow tie\u2014standing in front of the reflective sculpture. Their formal clothing matches the description, and no elements contradict this core requirement.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a couple dressed in formal attire, bride in a white wedding gown and a man in a black suit. a black tie.standing in front of a large sculpture. The attire attire align the description of and the additional contradict this core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..45593dcbc743b4d58768b2c19b87c98dd4f60a5b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:183b4c905628c749790fa8ae32f50f3818ba579f76d33b9bfaf988ce0b3ed87f
+size 172713
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..f33e4bb628c6d48e3f29c1c8427b15504fb346ec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a football player in a black jersey with the number 28 on it, wearing yellow gloves. He is standing on a football field with a blue wall in the background. The player is looking to his left with a focused expression. The style of the video is a sports action shot, capturing the player in a moment of concentration before a play. The lighting is bright, suggesting it is daytime. The player's stance and the setting suggest that he is preparing for a game or practice."
+    ],
+    "video_ids": [
+        "0_R_9L63gHs_56_0to128"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A football player in a black jersey with number 28, wearing yellow gloves.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a black jersey with the number 28 and yellow gloves, which matches the description. The player's attire and equipment are clearly visible and consistent with the specified condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person player wearing a black jersey with the number 28 and yellow gloves. which matches the description provided The player is attire and equipment are consistent visible, consistent with the given condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cefa13c5714d0c7a0b46ee948a8a47f62afa18ac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:536d401922424db0aafcffa5970e8d0034f8ac49576a6213e3e77ec1fa4dda8f
+size 172610
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..2d2b691ede50821409af7f0133b8c3f87827a2fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a beautifully set dining table with a gold tablecloth and purple chairs. The table is adorned with plates, vases of flowers, and candles, creating an elegant and inviting atmosphere. The table is surrounded by a white rug, and a green curtain can be seen in the background. The video captures the essence of a sophisticated dining experience, with attention to detail in the table setting and the overall ambiance."
+    ],
+    "video_ids": [
+        "tdPtCTc5_uU_1_0to117"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Gold tablecloth, purple chairs, plates, vases of flowers, candles? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a gold tablecloth covering the table, two purple upholstered chairs flanking the table, plates set for dining, vases containing white flowers, and tall candles in glass holders. All elements described in the condition are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a table tablecloth, a table, purple purple chairsstered chairs,anking the table, plates set on a, aases of flowers and, and candles candles in orange holders. The these match in the condition are present and match depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fb5064d66704d0dae0eef23302b38bf620be1f6b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f464441e6330e96c1ba2509778e3144311ba21c77a8c5e0171d5c78cccb75cd
+size 190168
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..eb6637f72d4d3da8aeed970a6cdb20e4aba38a66
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a group of six astronauts posing for a photo in a space shuttle. They are all wearing blue and red shirts with a logo on the left chest. The astronauts are smiling and appear to be in good spirits. The space shuttle is visible in the background, with the American flag prominently displayed. The lighting in the video is bright and even, suggesting that the photo was taken during the day. The astronauts are arranged in a pyramid formation, with the tallest astronaut at the top and the shortest at the bottom. The overall style of the video is professional and polished, with a focus on capturing the astronauts in a positive and confident light."
+    ],
+    "video_ids": [
+        "U3re0A_Piec_78_0to110"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Six astronauts posing for a photo, wearing blue and red shirts with a logo on the left chest.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows six astronauts posing together, wearing either blue or red shirts, each with a logo on the left chest. The background includes an American flag and what appears to be the interior of a spacecraft, which is consistent with an astronaut photo opportunity. The description is largely accurate, and no elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a individuals posing for, wearing blue blue or red shirts with which with a logo on the left chest. The presence includes an American flag, a appears to be a exterior of a spacecraft, which align consistent with the astronaut setting shoot. The description align largely fulfilled, with the significant contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..733bfe5d76df4dae98a9f94434940f43ae44459b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1321-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:33400bb83a69f05fce153a9d6ca80bffb92a2f515e44c2151f49e6bd06d29646
+size 170371
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..55e0cf37f0fd02b72298412757d1c8f47bd3124c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of celebration and camaraderie among a group of hockey players. The players, dressed in their vibrant red and blue jerseys, are seen high-fiving each other in front of a glass partition. Their joyous expressions and energetic gestures suggest a recent victory or a successful play. The glass partition, adorned with advertisements, separates them from the cheering crowd in the background. The crowd, a mix of individuals, is seen clapping and cheering, their faces filled with excitement and enthusiasm. The overall atmosphere is one of triumph and shared happiness, a snapshot of a moment that encapsulates the spirit of teamwork and sportsmanship."
+    ],
+    "video_ids": [
+        "kpzSBaSjrbI_4_41to202"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Hockey players, glass partition, cheering crowd? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing hockey players in red jerseys, a glass partition separating them from the audience, and a cheering crowd in the background. All specified elements are present and accurately depicted without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as depicting showing hockey players in red jerseys, a glass partition separating them from the cheering, and a cheering crowd in the background. The elements elements are present and do depicted.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6141cb156b33502a376a3f50cababc267cf917f4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0cd1d80d25dbd5d3c52740b8178e88f71be6f4588b75a89693effc418c3fb020
+size 361148
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..e2b84ce71ac3a49396018cba3f0805f2b44eabe4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene underwater scene featuring a clear, shallow body of water teeming with aquatic life. The water is dotted with small, delicate pink flowers that grow from slender stems, creating a vibrant and colorful environment. Several fish swim gracefully through the water, their movements fluid and unhurried. Among them, a koi fish stands out with its distinctive orange and white pattern, adding a splash of color to the scene. The fish move in various directions, some swimming close to the bottom where the sandy substrate is visible, while others glide near the surface. The overall atmosphere is tranquil, with the gentle flow of water and the graceful movements of the fish contributing to a peaceful ambiance. The camera remains stationary throughout the sequence, allowing viewers to fully absorb the beauty and tranquility of this underwater world."
+    ],
+    "video_ids": [
+        "8d1858d87e59e364ae182c68768dea1219281d46ea2f29c82a68363158c55c0a"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Fish, koi fish? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows multiple fish, including koi fish, swimming in clear water with aquatic plants. The koi fish are identifiable by their colorful patterns (white, orange, black), which matches the description. The presence of other fish does not contradict the core requirement of showing fish and koi fish.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts fish fish swimming including whatoi fish, swimming among an water among pink plants. The presenceoi fish are distinct by their distinctive patterns,orange, orange, and, which are the description of The presence of other fish and not contradict the ' description of the fish, koi fish.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4b111fb2ddea8b3b81b861845e2aa8eebccd9a49
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b9b044cf656a06bc3bf1e77f33384b0237cd98be2296238455ae643f079f801
+size 315576
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..d6a9253127913abfc55957bf618b0a9354ffa342
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen preparing a smoothie in a kitchen. The individual is pouring a creamy liquid, possibly yogurt or milk, into a blender filled with various ingredients. The blender contains a mix of green and brown ingredients, suggesting a healthy and nutritious blend. The kitchen setting includes wooden cabinets and a sink, indicating a home environment. The person is wearing a black apron, suggesting they are in the middle of cooking or food preparation. The overall style of the video is casual and homey, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "GnbVb9Pb448_6_0to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person in a black apron is pouring a creamy liquid into a blender containing green and brown ingredients.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person wearing a black apron over a red shirt, pouring a creamy liquid (likely peanut butter or a similar sauce) from a small glass container into a blender. The blender already contains visible green leafy ingredients and brown beans, matching the description. The action and objects are consistent with the stated condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person in a black apron pouring a white shirt, pouring a creamy liquid intolikely milk butter) a similar substance) into a white white container into a blender. The blender already contains green green andy ingredients, possibly chunks, which the description of The setting and setting in consistent with the given condition.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..908d310d269e91015743f9e3fb96e9d75c77a5f7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7832f64737eef151a01e64b8fa4df3871a521049c1167af887ca54dcdb6872b0
+size 136283
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..e278b19fe2e1953c9e0ef9d508e41983b44cb1b7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene winter scene in a rural area. The landscape is dominated by a large field covered in snow, with a few patches of brown earth visible. In the distance, a red barn stands out against the white snow, adding a pop of color to the scene. The sky above is a clear blue, suggesting a sunny day. The overall style of the video is calm and peaceful, with a focus on the natural beauty of the winter landscape. The camera seems to be positioned at a high angle, providing a panoramic view of the field and the barn. The video does not contain any people or animals, allowing the viewer to fully appreciate the tranquility of the scene."
+    ],
+    "video_ids": [
+        "SoKXA4e_T_s_30_0to192"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red barn in the distance.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red barn in the distance, situated among fields and trees, which matches the described condition. The barn is visible throughout the video, and no elements contradict its presence.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a red barn in the distance, which on a covered surrounded, which align the description '. The presence is the and the frames, and there other contradict the presence.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..764ded683f4747729f1366f92c2036bc9f200bb3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ebddd7bfec022d7c2adc14bfddf8cbf0a2e0478a8cd9be8f19ba54a2f7432f12
+size 150679
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..5edf197255a82b92fc3b5be1fd180dbf21194bb5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman in a kitchen, smiling and holding a wooden spoon. She is wearing a light blue denim dress and has her hair styled in braids. The kitchen has white cabinets and a backsplash with a geometric pattern. There is a stove in the background, and a bowl and a cutting board are visible on the counter. The woman appears to be cooking or preparing a meal, and the overall atmosphere of the video is cheerful and inviting."
+    ],
+    "video_ids": [
+        "5Ah_EM2CbV4_15_28to151"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a wooden spoon, a light blue denim dress, braided hair, a stove, a bowl, and a cutting board.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with braided hair wearing a light blue denim dress, holding a whisk (not a wooden spoon, but a similar kitchen tool). She is in a kitchen with a stove visible in the background. A cutting board and a stack of bowls are also visible on the counter. While the object 'wooden spoon' is not present, the other listed objects are present, and the video largely matches the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a woman wearing braided hair wearing a light blue denim dress, stirring a wooden,which a wooden spoon as but it similar utens utens). She is in a kitchen with a stove, in the background. There black board and a bowl of bowls are also present, the counter. The the wooden 'wooden spoon' is not present, the description elements elements are accurately and and the scene largely matches the description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..471f38a60792b931e910a2bcea4e9c95e8c3ca45
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f64a7e98ee36980e4c9f214817d2c2ba78a323213b0da37bacda4e63e07d5876
+size 149632
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..c9da53a4c06d881e78844dfea869a3f6f519a984
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red SUV parked in a parking lot. The car is parked next to other cars, and there are trees in the background. The car has a yellow license plate with the letters \"FGI50KS\". The car is parked in a parking space, and there is a fence in the background. The car is the main focus of the video, and it is parked in a stationary position. The other cars in the background are parked in a similar manner. The trees in the background add a touch of nature to the scene. The fence in the background provides a boundary for the parking lot. The license plate on the car is clearly visible, and it is a yellow color. The car is parked in a parking space, and there is a fence in the background. The car is the main focus of the video, and it is parked in a stationary position. The other cars in the background are parked in a similar manner. The trees in the background add a touch of nature to the scene. The fence in the background provides a boundary for the parking lot. The license plate on the car is clearly visible, and it is a yellow color. The car is parked in a parking space, and there is a fence in the background. The car is the main focus of the video, and it is parked in a stationary position. The other cars in the background are parked in a similar manner. The trees in the background add a touch of nature to the scene."
+    ],
+    "video_ids": [
+        "oKQEJXWebjw_8_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red SUV with a yellow license plate 'FGI50KS'.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red SUV, which is clearly identifiable as a Skoda Yeti, and the yellow license plate 'FGI50KS' is visible on the rear of the vehicle. These elements match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red SUV with which matches consistent visible as the Tataoda modeli, a it license license plate readsTI50KS' is visible. the front of the vehicle. The elements match the description provided, the questionObject(s)' condition.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e80e5ff3cf89e0f13447ef5e56a335303c6ec994
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34d0793e85d39b5038bc03ecaab3b275cf6c7971a531c222e93faa7f6f9f85bf
+size 278373
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..fec8303db5cab94bd0c3dc06f24c21bb7cfd64d3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a cooking tutorial, featuring a wooden cutting board with various ingredients and a glass bowl. The ingredients include sweet potatoes, green peppers, and cherry tomatoes. The style of the video is simple and straightforward, focusing on the preparation of a healthy and nutritious meal. The cutting board is the main subject of the video, with the ingredients arranged neatly around it. The glass bowl is used to hold the sweet potatoes, and the green peppers and cherry tomatoes are placed on the cutting board. The video is likely to be informative and educational, providing viewers with a step-by-step guide to preparing a delicious and healthy meal."
+    ],
+    "video_ids": [
+        "hHTCykx101w_5_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Wooden cutting board, sweet potatoes, green peppers, cherry tomatoes, glass bowl? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a wooden cutting board with sweet potatoes, green peppers, cherry tomatoes, and a glass bowl as central elements. These items are prominently displayed and match the described objects without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a wooden cutting board with sweet potatoes, green peppers, cherry tomatoes, and a glass bowl. the elements. The objects match present displayed, match the description objects. any or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e929e3898efa6d957d6a54ccf6383a162c94e4ee
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbcf3844ea1e90ed86a82950fb8db06ea52cefe00aed7b7b982e83df636dc186
+size 62478
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..0f63e1139bfe29c61cabf8fd5571ffebaca48c3a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a person enjoying a meal at a restaurant. The meal consists of a piece of food, possibly a dumpling or a small appetizer, served on a green leaf. The person is using chopsticks to pick up the food, indicating that the cuisine might be Asian. The table setting includes a wine glass, a fork, and a spoon, suggesting a mix of Western and Eastern dining styles. The background features a red tablecloth with a floral pattern, adding a touch of elegance to the scene. The overall style of the video is casual and intimate, focusing on the simple pleasure of enjoying a meal."
+    ],
+    "video_ids": [
+        "GADWWrm8VdY_934_36to172"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person using chopsticks, a piece of food (dumpling or small appetizer), a green leaf, a wine glass, a fork, and a spoon.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person using chopsticks to pick up a piece of food (which appears to be a dumpling or small appetizer) from a green leaf on a plate. A wine glass, fork, and spoon are also visible on the table, matching the described elements. The scene is consistent with the requested objects without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person using chopsticks to handle up a piece of food,which appears to be a dumpling or small appetizer) from a green leaf. a plate. There wine glass, a, and spoon are also visible on the table, fulfilling the description objects. The presence is set with the given objects and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..044b1b17302e7eedc815a786957fa1bc5d0a2f0a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1322-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e213e9a89f9f5d485882a5b86318930b57657e5086ac529637953f7dc04f2452
+size 128593
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..874ddaae19bec0eef37ecab8378408a71c37bfb9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a close-up view of a plate of seafood pasta, featuring linguine noodles mixed with shrimp, squid, and chunks of chicken. The dish is garnished with slices of tomato and sprinkled with herbs, giving it a vibrant and appetizing appearance. The pasta is coated in a light sauce, likely a garlic or herb-based dressing, enhancing its visual appeal. In the background, there is a glass of dark-colored beverage, possibly soda or iced tea, placed on a textured surface that resembles a tablecloth with a blue and white pattern. A fork is also visible on the left side of the frame, indicating readiness for eating. The scene remains static throughout the video, focusing on the delicious meal without any noticeable changes or movements."
+    ],
+    "video_ids": [
+        "e2a5be1d8b8b50f6ee57abbee80aa4fd393a9deb325bcdb2fafffd4cbfed87e5"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: seafood pasta (linguine noodles, shrimp, squid, chicken chunks), tomato slices, herbs, light sauce, glass of dark-colored beverage, fork? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a seafood pasta dish with linguine noodles, shrimp, squid, and chicken chunks, accompanied by tomato slices, herbs, and a light sauce. A glass of dark-colored beverage and a fork are also visible, matching the description. There are no conflicting elements in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showcasing showing a plate pasta dish with linguine noodles, shrimp, squid, and tomato chunks. along by tomato slices and herbs, and a light sauce. Additionally glass of dark-colored beverage and a fork are also visible in which the description provided The are no contradictions elements in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..481d7ce7a5d0c290655931251380e49d70078af8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d0ba5b4f54cd65c293ede3a49510f28ec4fc2fa1cb6d3d73ea976c0a6e7ee799
+size 79532
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..e10fa6af0f228040ce3ea4f632547e54cbe198e1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a silver SOTA wheel in three different stages of rotation. The wheel is designed with a black center and silver spokes, giving it a sleek and modern appearance. The SOTA logo is prominently displayed on the center of the wheel. The wheel is shown in motion, with each frame capturing a different angle of the wheel's rotation. The style of the video is a time-lapse, showcasing the wheel's design and movement in a dynamic and engaging way."
+    ],
+    "video_ids": [
+        "PDZL0lXUL0o_4_64to207"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Silver SOTA wheel? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a silver SOTA wheel with the brand logo clearly visible in the center. The wheel's design, including its spokes and bolts, is consistent with the description. No conflicting elements are present that contradict the core description of a silver SOTA wheel.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a silver wheelOTA wheel, a logo name ' visible in the center. The wheel's design and including its multiple and the, is consistent with the description of The additional elements are present, would the core description of the silver SOTA wheel.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1252029b7e4959d8995b4e24f37ed641840a9dca
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5be93bebe7db80b793cba678f47ba537813d90d1a717a54c1df438f2c9ab34ba
+size 190844
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..342675c2231f864bf02b501464ae7a529a81da1c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman named Vivian Chan, as indicated by the text overlay. She is wearing a colorful, patterned shirt with a star design and a matching apron. The setting appears to be a kitchen or a food-related environment, as suggested by the presence of a refrigerator and a shelf in the background. Vivian is speaking, as indicated by her open mouth and the microphone she is wearing. The style of the video seems to be a casual, personal vlog or a cooking tutorial, given the context and the woman's attire."
+    ],
+    "video_ids": [
+        "K3Bu5IAyW80_1_61to261"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman named Vivian Chan, a colorful patterned shirt with a star design, a matching apron, a microphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman who is identified as Vivian Chan via on-screen text. She is wearing a colorful patterned shirt with star designs and a matching apron. A microphone is visible clipped to her shirt. All elements of the description are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing appears wearing as Vivian Chan, the-screen text. She is wearing a colorful patterned shirt with a designs and a matching apron. There microphone is visible in to her shirt, The elements of the description are present and match with the video content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2c793bab220e5f61dbddc84f7434ce7ecd303e29
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0d6bcc0d90f091a58cf696d3cb6256fa83e7379c5c4e6a0e4f3132e4063d0d85
+size 270983
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..a08f03c182042b4aafea4542daefbb6a8a2cb8f3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a pumpkin with a carved face, featuring a wide, toothy grin and two eyes. The pumpkin is placed on a white surface, and the camera angle is slightly elevated, looking down at the pumpkin. The style of the video is simple and straightforward, focusing on the pumpkin as the main subject. The lighting is bright, highlighting the details of the carved face. The background is plain and unobtrusive, ensuring that the viewer's attention remains on the pumpkin. The video does not contain any additional objects or actions, and the pumpkin remains stationary throughout the frames. The overall impression is of a traditional Halloween decoration, with a focus on the craftsmanship of the carved face."
+    ],
+    "video_ids": [
+        "K0GxmnFdjfE_16_0to192"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A pumpkin with a carved face featuring a wide, toothy grin and two eyes.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a pumpkin with a carved face that includes a wide, toothy grin and two eyes, matching the description. The additional element of a small pumpkin carved as a pig's snout in the mouth does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a pumpkin with a carved face that features a wide, toothy grin and two eyes. which the description provided The pumpkin elements of the white, on with a facelet headout does the background does not contradict the main description but}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..17298f1f0179e6490c16dfc922f49afe69a54ac9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1c44c2f1b2267d9de9ba194acb643cdf7140f8de3340d4a9d75848d4b32e6794
+size 96052
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..614719834a995c5d5affb6b742af8fe8b5a9a7f4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in front of a blue curtain. He is wearing a blue shirt and has a beard. His hands are raised, and he appears to be speaking or gesturing. The style of the video is casual and informal, with the man being the main focus. The blue curtain provides a simple and uncluttered background, allowing the viewer to focus on the man and his actions. The lighting in the video is bright, suggesting that it was filmed during the day. The overall impression is that of a friendly and approachable individual, possibly engaging in a conversation or presentation."
+    ],
+    "video_ids": [
+        "1hSrC227IF8_0_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a blue curtain? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man standing in front of a blue curtain, which matches the 'Object(s)' condition. The man is the central subject, and the blue curtain serves as the background, fulfilling the description without any conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features shows a man standing in front of a blue curtain. which ful the 'Object(s)' condition. The man is the main figure, and the blue curtain serves as the background, fulfilling the description provided any contradictions elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0aed957a3f7252cd7b741400ce3b37603a9c4972
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8a782846daeaba6edabd716a0f90254afdae6f2c4c6cbad89489d8e78bd4eb76
+size 202625
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..08bd5870bad706506f3174a7eec93b6524837b94
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse of a car's interior, showcasing the dashboard and steering wheel. The car is a Mercedes-Benz, as indicated by the logo on the steering wheel. The dashboard is sleek and modern, with a digital display that shows the car's speed and other information. The steering wheel is black and has a red center, adding a touch of color to the otherwise monochrome interior. The car's interior is well-lit, with natural light coming in through the windows. The seats are black and appear to be made of leather. The car is parked in a lot, as evidenced by the other cars visible in the background. The video does not show any people or animals. The style of the video is straightforward and unembellished, focusing solely on the car's interior."
+    ],
+    "video_ids": [
+        "Y_QGqWUsxqg_30_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Mercedes-Benz car, dashboard, steering wheel, seats? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the interior of a Mercedes-Benz car, including the dashboard, steering wheel, and seats. All specified objects are prominently featured and accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows the interior of a Mercedes-Benz car, including the dashboard, steering wheel, and seats. The the objects are present visible, match depicted, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e75b6263852894e31b98475f08396ea516cca930
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:225355a08ea98116acecfe52d279e6bfed6dc408ffe29e12386eceffcb1a964c
+size 207494
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..a58666e92bfedaf38d57b9f2dc0374f5dea8a67e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a group of people walking on a circular glass bridge that extends over a deep canyon. The bridge is suspended high above the canyon floor, providing a thrilling view of the vast, rocky terrain below. The group, dressed in casual attire, moves along the bridge, some looking down at the canyon while others gaze out at the surrounding landscape. The bridge's design allows for a clear, unobstructed view of the canyon, creating a sense of awe and wonder. The video is shot from a high angle, emphasizing the height and daring of the bridge and the people traversing it. The overall style of the video is adventurous and exciting, capturing the thrill of exploring nature's grandeur from an unusual vantage point."
+    ],
+    "video_ids": [
+        "IkJNeLo7XTI_25_0to134"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A group of people? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a group of people walking and standing on a glass bridge overlooking a canyon. Their presence and activity match the description of 'A group of people' without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a group of people walking on standing on a glass bridge. a canyon. The presence and the fulfill the description of aA group of people'. in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..940b64f094af67a9c487d8dc889cc5e0b0d7f0e2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:661bfef7b361c401ed604e2e8cc5b3b1b1b55d98877bac1c63eeed870e9b404c
+size 315942
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..3ea9f6db84feade56b20b83f71c46388fc29be91
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a large, intricately designed mosaic floor situated on a rocky beach. The mosaic features a geometric pattern with a variety of colors, including shades of red, black, and white. The floor is surrounded by large, jagged rocks, and the ocean can be seen in the background, with waves crashing against the shore. The sky is clear and blue, suggesting a sunny day. The style of the video is realistic, with a focus on the natural beauty of the location and the detailed craftsmanship of the mosaic. The camera angle is low, looking up at the mosaic, which emphasizes its size and complexity. The overall mood of the video is serene and peaceful, with the sound of the waves and the sight of the ocean creating a sense of tranquility."
+    ],
+    "video_ids": [
+        "U4WZndFzC7w_3_0to152"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Large, intricate mosaic floor, jagged rocks, ocean, crashing waves, clear blue sky? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a large, intricate mosaic floor in the foreground, jagged reddish-brown rocks along the shoreline, an ocean with crashing waves, and a clear blue sky with some clouds. All elements described in the condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a large, intricate mosaic floor with the foreground, withged rocksish-brown rocks in the shoreline, the ocean with crashing waves, and a clear blue sky. a clouds. The elements are in the condition are present and contribute represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b9257e0e5f73b2cce2668a54abd63293f7de68fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1323-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c32b109dbe7b78d560c3bbb684cbc821cb795af826058aaa3ff7922cdc3eba2b
+size 196490
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..4e506d59d42bab208b84820dae6c906cf0a4b03e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a luxury car, focusing on the driver's side. The car's interior is sleek and modern, with a black color scheme. The steering wheel is in the center of the frame, with the dashboard and center console visible. The car's door is open, revealing the door handle and window controls. The car's interior is well-lit, with the sunroof open, allowing natural light to fill the cabin. The car's design and features suggest it is a high-end model, likely a sports car or luxury sedan. The video captures the car's interior in a way that highlights its design and features, providing a glimpse into the experience of driving this luxury vehicle."
+    ],
+    "video_ids": [
+        "6vq1kH65dNI_53_0to157"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, dashboard, center console, door handle, window controls, sunroof.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the steering wheel, dashboard, center console, door handle, and window controls. The sunroof is not visible, but since the video is taken from an interior perspective with the roof open, it is implied that the car has a convertible top rather than a fixed sunroof. The absence of a visible sunroof does not contradict the core description, as the video still fulfills the 'Object(s)' condition for the other listed items.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows the interior wheel, dashboard, center console, door handle, and window controls. The presenceroof is not visible in but the the question is focused from the interior perspective, the door partially, it can reasonable that the sun has a sun top, than a sun sunroof. The presence of a sun sunroof does not contradict the description description of as the video focuses fulfills the 'Object(s)' condition by the other elements elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..978ed824e4bf1d3b84f86fc1cc67267c03e27831
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b503f1a4d0a5426bb8274a8847fd9257f404742ad3ab0b65664715845961e263
+size 214023
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..941b2aa8fbdb3aae5aa6ce8b7e9a1bb560d8a7f9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a suit standing at a press conference. He is speaking into a microphone, which is placed on a table in front of him. Behind him, there are two football helmets, one with the letter \"K\" and the other with the letter \"U\". The man appears to be addressing the media, possibly discussing a football game or event. The setting suggests a formal and professional atmosphere."
+    ],
+    "video_ids": [
+        "3Eax4umGSX8_2_30to195"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit, a microphone, two football helmets.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a suit seated at a table with a microphone in front of him. Two football helmets are also visible on the table, one on each side of the man. The background features the Big 12 Conference logo, which is consistent with the context of a press conference or media event. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man in a suit, at a desk with a microphone in front of him. There football helmets are visible visible in the table behind one in the side of the man. The presence is additional helmets Ten12 conference logo, which is not with the context of football sports conference or sports event related The elements elements of in present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c8f6aec0697a44702de6d6067b5de42e9bea6235
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:58d2881a8e4cc6f50301a3803dd3de34c5411d7dbe23a68180a416cb208c8f98
+size 184160
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..56458a3e621b1cbb9872bff1f09889b235099ac6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the growth of a tree with green leaves and small, round fruits. The fruits start as small buds, then grow into larger fruits, and eventually mature into fully formed fruits. The tree is lush and green, with the fruits hanging from the branches. The fruits are round and green, and they grow in clusters. The tree is in full bloom, with the leaves and fruits providing a vibrant contrast to the sky. The video is a time-lapse, showing the progression of the tree and its fruits over time. The style of the video is naturalistic, capturing the beauty of nature in a realistic and unfiltered way."
+    ],
+    "video_ids": [
+        "SFoZ1RtAmw8_6_60to188"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Tree, green leaves, small round fruits (buds), larger round fruits, fully formed fruits, sky? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a tree with green leaves and various stages of fruits: small round buds, larger round fruits, and fully formed fruits. The background is blurred but appears to be sky or distant scenery, which aligns with the description. No conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a tree with green leaves and small stages of fruits, small round fruits, larger round fruits, and fully formed fruits. The sky is a but appears to be a, a clouds, which aligns with the '. The additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..43bfb3a9ba3475492c6bd020b8b99c274225ee62
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7118ef211f7fe7ab09a009048a7b84d0e99a170675ead2aacea045b6b6a58742
+size 215511
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..6f1ef80939540903bf37637aabaefe577a9276a1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a bird's eye view of a coastal city, showcasing its natural beauty and urban development. The first frame captures the city's skyline, with buildings of various sizes and shapes, indicating a mix of residential and commercial areas. The second frame reveals a large body of water, possibly a bay or a river, with a bridge connecting two parts of the city. The third frame provides a closer look at the city's green spaces, with lush trees and vegetation covering the hills and mountains that surround the city. The video style is aerial, providing a comprehensive view of the city's layout and its relationship with the natural environment. The colors are vibrant, with the blue of the water contrasting with the green of the vegetation and the gray of the buildings. The video is likely to be used for promotional purposes, highlighting the city's unique features and its potential for tourism and investment."
+    ],
+    "video_ids": [
+        "8T_PXRG4apc_45_42to224"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Buildings, a large body of water, a bridge, lush trees and vegetation, hills and mountains? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a large body of water (the sea), lush trees and vegetation covering rolling hills and mountains, and a distant urban area with buildings. While a bridge is not distinctly visible, the long structure extending into the water may be interpreted as a bridge or causeway, which aligns with the description. The overall scene matches the requested elements without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as showcasing showing buildings large body of water,a lake), a trees and vegetation ( the hills and mountains, a a bridge bridge area with buildings. Additionally the bridge is not explicitly visible in the presence structure in over the water could be interpreted as a bridge, away. which iss with the presence. The presence scene includes the provided elements without any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0002aa1784536f88d14aee0e65e7d716df10dee7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dbec4f22ae5c327aca6169e0fa43dfd97b666ecad8569fd5422d7d6d7d3dce30
+size 258545
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..665f3e7caafa62e19d6ad0c50c128bd969842d34
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman standing in front of a blue background, speaking into a microphone. She is wearing a pink dress and has long brown hair. On either side of her, there are two images of lions. The lions are depicted in a realistic style, with one on the left and one on the right. The woman appears to be presenting or discussing something, possibly related to the lions or the DNA sequence shown in the background. The overall style of the video is informative and educational, with a focus on the woman's speech and the images of the lions."
+    ],
+    "video_ids": [
+        "UPcT3qOs2Hk_1_21to151"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, two images of lions, a microphone, and a DNA sequence image.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a woman, multiple images of lions (which increases beyond two as the video progresses), a microphone (visible as a small device on her collar), and a DNA sequence image (a prominent double helix on the left side of the screen). All core elements are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video contains fulfills the 'Object(s)' condition as featuring showing a woman holding two images of lions,two appear the the), there video progresses), a microphone,held in she single object in the left), and a DNA sequence image (part white white helix structure the right side of the background). The these elements are present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c30ae58713905a052990a40f8aa7c5223f60716b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:35f836973d0099aa508740184097b98e84bf7b58d25f4a8f764468f6816c794e
+size 130731
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..5cc4861861155bd28158302bc050791ca61f2fe0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a kitchen, preparing a drink. She is wearing a pink shirt and is standing in front of a window. The kitchen is well-lit and has a clean, modern look. On the counter, there are several items including a cutting board, a knife, and a glass of water. The woman is seen cutting lemons and placing them into the glass of water. The overall style of the video is casual and homey, with a focus on the woman's actions and the bright, inviting kitchen setting."
+    ],
+    "video_ids": [
+        "WYX4uCF1YOE_8_39to174"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a cutting board, a knife, a glass of water, lemons? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman standing at a kitchen counter. In front of her is a wooden cutting board with sliced lemons and a knife. There is also a glass pitcher filled with water. All the specified objects are present and visible in the scene, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a woman in at a kitchen counter, She front of her, a cutting cutting board where a lemons on a knife. There is a a glass of of with water on The the elements objects are present and the in the scene, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..90c667218f10f3d2fac4b6d1776581aeb1c7f70c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:44a60b73617d5e6fac2d697cd7d722b63ac29df892dac2da899d783f81b1b77e
+size 127482
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..fb073e922c40ee200d561ff4a38fd13ff7ad2db4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in a snowy landscape, wearing sunglasses and a black jacket. He is positioned in front of a car, which is parked on a snow-covered road. The man appears to be speaking or gesturing towards the car, possibly discussing its features or performance. The background of the video shows a serene winter scene with trees and a body of water, suggesting that the location is a scenic, possibly remote, area. The overall style of the video is casual and informative, with a focus on the man and the car in the context of the winter environment."
+    ],
+    "video_ids": [
+        "78oMgfeoAfw_23_0to174"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a car, trees, and a body of water? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing sunglasses and a black jacket, standing next to a car in a snowy environment. In the background, there are bare trees and a body of water, which aligns with the specified objects. The scene is consistent with the description without any contradictory elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a man standing sunglasses and a black jacket, standing next to a silver in a snowy environment. There the background, there are trees trees and a body of water, which iss with the description objects. The presence is consistent with the description provided any contradictions elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8b660f706b8420112ef509ea7376b14370451301
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1229360cc4201d86015d7f866a80eaeb45dd2905d64d33446d7609d933340732
+size 251573
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..3779ee3454c6996b3c7fed74a1744808d69c4cfc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit and tie, sitting in front of a blue background with a white door. He is wearing glasses and has a microphone attached to his ear. The man appears to be speaking or listening intently. The video has a news-like style, with a logo in the corner that reads \"El Rumbo de la Econom\u00eda\" and \"Lo leu\". The man is the main subject of the video, and the focus is on his face and upper body. The background is simple and does not distract from the man. The overall style of the video is professional and straightforward, typical of a news segment or interview."
+    ],
+    "video_ids": [
+        "HDu8NMzFxxo_2_251to421"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, wearing glasses, with a microphone attached to his ear; a logo in the corner reading 'El Rumbo de la Econom\u00eda' and 'Lo leu'.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit and tie, wearing glasses, with a microphone attached to his ear, which matches the description. The text overlay includes 'EL RUMBO DE LA ECONOM\u00cdA' and 'los leuco', which aligns with the logo mentioned. There are no conflicting elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a man in a suit and tie, wearing glasses, which a microphone attached to his ear, which align the description. The logo in in aEl RUMBO DE LA ECONOM\u00cdA' and 'LO leu', which iss with the ' mentioned in The are no contradictions elements in contradict the description description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f28296ea1b2cf0daeb261cfff7d6014ed6b4ad43
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1324-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd874ee0f55e76457e8e7632c28dfa178e73ae93fab925e4dc8c18c4c34b6131
+size 157054
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..682d0628d84774ca39a543f88aa32323f85218fe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a luxury car, showcasing its sleek design and advanced technology. The car's dashboard is adorned with a large touch screen display, which is the focal point of the video. The steering wheel, featuring the car's logo, is positioned in the center of the frame, drawing attention to the car's brand. The car's interior is upholstered in black leather, adding to its luxurious feel. The video is shot from the perspective of the driver, providing a comprehensive view of the car's interior. The car's design and technology are highlighted, making it an appealing choice for those seeking a high-end driving experience."
+    ],
+    "video_ids": [
+        "tZ5UXmWtpuk_48_0to123"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dashboard, touch screen display, steering wheel, car's logo, black leather upholstery? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows the dashboard, touch screen display, steering wheel, car's logo (Bentley emblem), and black leather upholstery as described. All these elements are clearly visible and accurately represented in the frames, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful the interior, touch screen display, steering wheel, car's logo,Merenzley),), and black leather upholstery. described. The these elements are clearly visible and match represented in the video provided fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4ffbf73562e19e7a213e50f02912fd05ff87f7d4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04eafee0f5dc931ff6f2fbc5e7c5074131c2ba4e2c1bed1fc3c2a5e33708a7cd
+size 198262
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..5570457e6a37c01781fa9ad979623b6c2e332d32
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a soccer match. The main focus is on two players, one in a white jersey and the other in a red jersey. The player in the white jersey is in motion, running towards the right side of the frame, while the player in the red jersey is in a defensive stance, ready to intercept. The background is a blur of the soccer field and the crowd, indicating the high-speed action of the game. The style of the video is a fast-paced, action-packed sports sequence, capturing the intensity and excitement of the match."
+    ],
+    "video_ids": [
+        "qL-LIWITt6A_9_69to213"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two players, one in white jersey, one in red jersey.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two players: one in a white jersey (with the number 17 visible) and another in a red jersey (with the number 4 visible). The description of the two players' jerseys is accurate, and no elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two players, one in a white jersey andnumber the number 10)) and one in a red jersey.with the number 1 visible). Both players of the ' players' jerseys matches accurately, and the additional contradict this core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..be9d1f636777b05d0e3479ca7019a6a08ba90b47
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac2235a7d7ae9c47a84d0e79224b6223540b29f08336f37022fbeb51f19f436d
+size 186339
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..e772f3aab7739a5879b6acbf8c3b997b7d73a869
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a group of children in a classroom setting, engaged in a shared activity. The children are of various ages and ethnicities, and they are all focused on something outside the frame of the image. The classroom is filled with natural light, suggesting it is daytime. The children are seated on the floor, indicating a casual and relaxed atmosphere. The style of the video is candid and unposed, capturing a moment of learning and engagement among the children."
+    ],
+    "video_ids": [
+        "FHiijJGLzmQ_18_0to153"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Children of various ages and ethnicities seated on the floor.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group of children of various ages and ethnicities seated on the floor, attentively looking upward. Their diverse appearances and seating arrangement match the description, with no conflicting elements present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group of children of various ages and ethnicities seated on the floor, engagedively looking at, The diverse clothing and the arrangement on the description provided indicating no conflicting elements present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2d400b0b8a7da26c6134270df5b8394e7a84eee8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:979946277d160088efaaf0354d74b7907add4cd5d3631be0bb0c74c7d6473bb3
+size 207312
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..37301c6cdd6254698e93b8bed3b1332a14210c16
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with curly hair and glasses, wearing a patterned shirt, sitting in a chair and gesturing with his hands. He appears to be engaged in a conversation or presentation. The setting is outdoors, with a building and a car visible in the background. The man's expressive body language suggests he is passionate or enthusiastic about the topic at hand. The overall style of the video is casual and informal, with a focus on the man's facial expressions and hand gestures."
+    ],
+    "video_ids": [
+        "HjcuTrnrMf4_28_30to206"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with curly hair and glasses, wearing a patterned shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with curly hair and glasses, wearing a patterned shirt, which matches the description. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a person with curly hair and glasses, wearing a patterned shirt. which matches the description provided The man includes additional elements in not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..eb983a01d13d0508485eb6acaa891ef37b4c876c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:74655325b5cdeb52e8c9cfaa5650484ed2bc6813c3ed87295298f3725a2339f4
+size 289128
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..5dd64097e2de05648d7432a263b2ef25775b66b9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man with a beard and glasses, wearing a green hoodie and a baseball cap, sitting in the driver's seat of a car. He is holding a smartphone in his right hand and appears to be looking at the screen. The car's dashboard is visible, with the radio and air conditioning controls. The man seems to be adjusting the settings on his phone, possibly navigating or changing the music. The car is parked, and the interior is well-lit, suggesting it's daytime. The style of the video is casual and candid, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "18BqKMkoLcU_36_0to153"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and glasses, wearing a green hoodie and a baseball cap, and a smartphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and glasses, wearing a green hoodie and a baseball cap, seated in a car. He is holding a smartphone in one hand and interacting with the car's infotainment system with the other. All core elements of the description are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and glasses, wearing a green hoodie and a baseball cap, sitting in a car. He is holding a smartphone and his hand and appears with it device's gearotainment system with the other. The these elements of the description are present and accurately depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0d89c577abc964e002c22c14119837d5f5aaf534
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:16ccb4474d40c3ed25a7c95af5398eb43a366beecbf8a4659c68a51ff77a86cf
+size 193627
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..c6a8c2d10a5edfa664294ac5528a599e0eb05a7f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a man in a red outfit standing and talking to another man who is seated in a chair. The standing man is holding a book and appears to be engaged in a conversation with the seated man. The setting is an outdoor patio area with wooden furniture, including chairs and tables. The patio is surrounded by plants and has a view of a street with parked cars. The lighting suggests it is daytime. The style of the video is casual and appears to be a real-life scenario rather than a staged or scripted event."
+    ],
+    "video_ids": [
+        "DaIiszLPRSU_17_48to169"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a red outfit, another man seated in a chair, and a book.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a red outfit standing and holding a book, while another man is seated in a chair. These core elements are present and accurately match the description. Additional elements like chairs, tables, and cars in the background do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man in a red outfit standing and holding a book, while another man is seated in a chair. The elements elements match present and match depicted the description. The elements such the, a, and a in the background do not contradict the main and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fbb492f7ccda8abeb0d553498359d528cf83a6a9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:adc51ff59fd72ccce6a02ddf880a74f8ee62ada58ad76573987462d1b460064f
+size 137627
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..cb5ec5b9199f5851d7a7e631376ea6eff8e9996c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of three red sweet potatoes with their tops cut off, revealing the creamy white flesh inside. The sweet potatoes are placed on a wooden surface, and each one is filled with a creamy white substance, possibly a sweet potato filling or a white sauce. The sweet potatoes are arranged in a triangular formation, with the camera angle focusing on the one in the front. The lighting in the video is soft and warm, highlighting the natural texture and color of the sweet potatoes. The style of the video is simple and straightforward, focusing on the food without any additional elements or distractions."
+    ],
+    "video_ids": [
+        "vfC1cBmgLV8_1_22to178"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three red sweet potatoes? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three red sweet potatoes that have been halved and filled with a white creamy substance, drizzled with a golden liquid, likely honey or syrup. The description 'Three red sweet potatoes' is accurately fulfilled as the main objects in the video are indeed three such potatoes, matching the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three red sweet potatoes, are been cutved and filled with a white creamy substance. likelyizzled with a light liquid. likely honey or syrup. The description ofThree red sweet potatoes' is fulfilled represented by the video objects in the video are indeed three red sweet. and the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bdf1d2c798b3f01dbb07dd81fa58505f633201d9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f749a07de86695f89e0e3381ccd6c6a9d40dd5b6fb739133d3d1e5e4586d232b
+size 72802
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..499e49f7faf893619635a6a5e47d215c473b841d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person eating a meal from a white plate. The plate contains a variety of food items, including grilled chicken, lettuce, and a side of sauce. The person is using a fork to pick up a piece of chicken from the plate. The style of the video is a simple, everyday scene, capturing a moment of someone enjoying their meal. The focus is on the food and the person's interaction with it, rather than any elaborate or artistic elements. The video does not contain any text or additional graphics."
+    ],
+    "video_ids": [
+        "AcBGd7ncvcc_16_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a white plate, grilled chicken, lettuce, sauce, fork? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person using a fork to eat from a white plate that contains grilled chicken, lettuce, sauce (guacamole and sour cream), and other toppings. All specified objects are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person using a fork to interact from a white plate that contains grilled chicken, lettuce, and,likelyacamole), possibly cream), and a vegetables. The the elements are present and the with the description.\"\n}<|im_end|>\nstyleType",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a41fadcfe9b5c2e1d5e37501ea6472e1f3b9b177
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1325-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7184da7d7160bbec60982a6f4245ccb863a29d3f9d924cebbe90f2b4e3a19024
+size 150886
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..b6dc182c87f46cded8267d70adc1ddc2c50605b5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features two animated characters, a dog and a cat, standing in a grassy yard in front of a house. The dog is on the left, wearing a blue shirt and khaki shorts, and is waving with its right paw. The cat is on the right, wearing a green shirt and blue pants, and is looking surprised. The house in the background has a red roof and white walls. The yard is green and well-maintained, with a white fence and a tree in the background. The sky is clear and blue, suggesting a sunny day. The characters appear to be in a friendly interaction, possibly greeting each other or preparing for a play. The overall style of the video is colorful and cartoonish, with a focus on the characters and their expressions."
+    ],
+    "video_ids": [
+        "R6pYlVXSivU_55_20to182"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two animated characters: a dog and a cat.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly features two animated characters: a dog (white with blue patches, wearing a blue shirt) and a cat (gray with green eyes, wearing a green shirt). These characters are the central focus of the scene, and their actions and interactions are the main content of the video, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts two animated characters: a dog andw with brown shirt) wearing a blue shirt and and a cat (orange with orange patches, wearing a green shirt). Both characters are standing main focus of the video, and there presence and expressions are the primary content of the video. fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..459990bc33a2f71e8c98a030d522f42fc23b9b44
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:deda7e757acc93528452531f7f40498bcde4c6dea9bd601c8dae30ef9f5421f6
+size 193262
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..6851766d815f5f11dfa59ef1246bb35983518bae
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person preparing a chocolate cake. The cake is placed on a silver platter, and the person is seen touching the top of the cake. The cake has a rich, dark brown color, indicating it is a chocolate cake. The person is wearing a white shirt, and the background is a kitchen counter. The style of the video is a simple, everyday cooking video, focusing on the preparation of the cake. The person is the main subject of the video, and the cake is the main object being prepared. The video does not contain any other significant objects or actions."
+    ],
+    "video_ids": [
+        "-sylw0wEldw_18_161to318"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person, chocolate cake? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person interacting with a chocolate cake, which is the core subject of the description. The person's hands are visible, and they are handling the cake on a rotating stand, consistent with the 'Object(s)' condition. Additional elements like a plate with crumbs or kitchen background do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a person interacting with a chocolate cake. which ful the core elements of the description. The person's hands are visible, and they are touching the cake, a rotating stand, which with the 'Person(s)' condition. The elements like the marble and a and a tools do not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6a3248ec35130f5e9d45f6666ab926721cf142c6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d719460875a171ae51e30885e0b583ade531b366ef4d1ec7e4b1239c95ed5e69
+size 115100
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..edf6c60ede783ea08c2015016dce00bda7f7528b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene set in what appears to be a tranquil garden or backyard. A cat, with a mix of brown and white fur, is perched on the edge of a small, circular pond. The pond is bordered by a black rim and contains clear water where a large, light-colored koi fish swims gracefully. The cat's attention is captivated by the fish, which moves slowly across the surface of the water. The background features a variety of plants, including shrubs and bushes, adding to the natural ambiance. The setting includes rocks and a stone structure, possibly part of a decorative fountain, enhancing the peaceful environment. The cat remains stationary, its gaze fixed on the fish, creating a moment of quiet interaction between the two animals. The overall atmosphere is calm and picturesque, highlighting the harmony between the cat and its surroundings."
+    ],
+    "video_ids": [
+        "5ff086176c10155ec9c33033c9cec4155b5a63a8b8da08b8ae86762ba9fe6f0c"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A cat with brown and white fur and a large, light-colored koi fish.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a cat with brown and white fur leaning over the edge of a pond, and a large, light-colored koi fish swimming near the surface. The core elements described in the condition are clearly present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a cat with brown and white fur, over a edge of a pond, observing a large, light-colored koi fish swimming in the cat. The cat elements of in the question are present present, match depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d12bb74f0b954cb430a5154640614b084c233269
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3516308298ada6356a170eeb7af3c08fecc9863401f1d61e2ec1ab7c11b0febc
+size 192589
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..481d7913767b785524724f3bdfbd85f33bb915ff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a chef in a professional kitchen, meticulously crafting chocolate truffles. The chef, dressed in a white chef's coat, is seen using a piping bag to fill the chocolate truffles with a rich, dark chocolate ganache. The truffles are neatly arranged on a white tray, ready to be filled. The chef's focus and precision are evident in the careful way he handles the piping bag, ensuring each truffle is filled to perfection. The kitchen setting is professional and well-equipped, with a wooden table and a sink visible in the background. The overall style of the video is clean and professional, capturing the artistry and skill involved in the creation of these chocolate truffles."
+    ],
+    "video_ids": [
+        "BzopOFwv28c_2_254to377"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chef, piping bag, chocolate truffles, white tray? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a chef in a white chef's coat using a piping bag to dispense chocolate onto a white tray, forming chocolate truffles. All the specified objects are present and accurately depicted in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a chef in a kitchen uniform's coat, a piping bag to decorateense chocolate onto a white tray. which chocolate truffles. The the elements elements ( present and the depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..62a5e31ca4742d981f227b724a2d6bdd31387403
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03d541135e01a5e7f7a50fd5635d8ad3367c40d6ff23dd0ffda7ddfdabe5a39a
+size 190882
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..175ee430cb9b0595ef588925f339b8f856364612
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is an aerial view of a cityscape, showcasing a large, modern building with a distinctive, curved, white roof. The building is surrounded by a large, blue, rectangular pool, which is filled with water. The pool is situated in the foreground of the image, with the building in the background. The cityscape is filled with various buildings, including skyscrapers and low-rise structures. The sky is clear and blue, suggesting a sunny day. The overall style of the video is a drone shot, providing a bird's eye view of the city and the building. The video captures the contrast between the modern architecture of the building and the surrounding cityscape."
+    ],
+    "video_ids": [
+        "1zcf_dgeFhs_2_173to324"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Large, modern building with a curved, white roof, large blue rectangular pool, various buildings including skyscrapers and low-rise structures? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large, modern building with a distinctive curved, white roof, surrounded by large blue rectangular pools. In the background, there are various buildings including both skyscrapers and low-rise structures, matching the description. The overall scene is consistent with the requested elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a large, modern building with a curved curved, white roof, which by a blue rectangular pools. The the background, there are various buildings, skys skyscrapers and low-rise structures, which the description provided The presence scene is consistent with the provided elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fbf701f04ea004090fb87b56c6040429bfc1c18f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e09dd6ecfca6f4547bd76549eb12b2bfc19dba840ca94bcdb320971204df58f5
+size 182061
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..83598bf2c22372b8a73d1223ec682e298b38c359
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person's hand reaching for a piece of fruit from a bowl filled with a variety of fruits, including strawberries, grapes, and slices of cheese. The bowl is placed on a white tablecloth, and there are two other bowls nearby, one containing crackers and the other containing olives. The style of the video is a close-up, slow-motion shot that captures the texture and colors of the food in detail. The focus is on the hand and the fruit, with the background being blurred to emphasize the action. The lighting is soft and natural, suggesting an indoor setting with natural light. The overall mood of the video is calm and inviting, highlighting the freshness and appeal of the food."
+    ],
+    "video_ids": [
+        "BNXlIMOA2Ag_10_0to140"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person's hand, piece of fruit (strawberries, grapes, slices of cheese), bowl with fruit, bowl with crackers, bowl with olives? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a person's hand, strawberries, grapes, slices of cheese, a bowl with fruit, a bowl with crackers, and a bowl with olives. All described elements are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showing showing a person's hand interacting a, grapes, slices of cheese, a bowl with fruit, a bowl with crackers, and a bowl with olives. The these elements are present and interact depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..30a790917650188621997772c09860f7ed052d46
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab74b3ee1d79f5bc1bcc3c102f1fd36191466131bcab32f8a116cd89a4eb2fc0
+size 172723
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..7b87cb9f1c11a9e324f12130155e8e2fde0569b6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a promotional advertisement for a hotel named \"Hotel Siete Lunas\". The style of the video is a combination of interior design and travel promotion. The video showcases the hotel's balcony, which offers a stunning view of the ocean and mountains. The balcony is furnished with a table and chairs, and a potted plant adds a touch of greenery. The interior of the hotel is decorated with wooden furniture and a rug, creating a warm and inviting atmosphere. The video is likely to be used on the hotel's website or social media platforms to attract potential guests."
+    ],
+    "video_ids": [
+        "_zg3GiXvl9k_21_0to199"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Balcony, table and chairs, potted plant, wooden furniture, rug? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a balcony with two wooden chairs and a small table, a potted plant on a wooden table, wooden furniture including a couch and ottoman, and a rug on the floor. All elements described in the condition are present and clearly visible.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a balcony with a chairs chairs and a table table, a potted plant on the wooden surface, and furniture, the chair and aoman, and a rug with the floor. The these match in the ' are present and match visible in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..081050504665333538cc09fd54911d73463df028
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0559436f8fdd22465c7f9932e3b2b39921e92261fefb3433a870948f1c426141
+size 143335
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..0f13a59f5df7cf557bca7ed6b31c376b88f9cf6f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a journey through a desert landscape, with a focus on a safety sign in the foreground. The sign, which is prominently displayed in the middle of the video, warns of potential dangers in the area. The background features a rocky cliff face, which adds a sense of ruggedness to the scene. The overall style of the video is realistic, with a focus on capturing the natural beauty of the desert landscape. The sign serves as a reminder of the potential dangers that can be encountered in such environments."
+    ],
+    "video_ids": [
+        "Uf0xE0vpZrk_9_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Safety sign in the foreground.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a safety sign in the foreground, which is the central object of focus. The sign is clearly visible throughout the video, and while there are other elements like rocks and vegetation in the background, they do not contradict the core description of a safety sign in the foreground.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a yellow sign in the foreground, which is yellow main object described the. The sign is yellow visible and the frames, and it the are additional elements like a and a in the background, they do not detr the description description of the safety sign in the foreground.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..575d8d05f9fbb5f34f6c382df192a9f356596557
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1326-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b4d326254e28302dbbbeb587f937e7982d82c9ebed7ccbcde1cc8f6611a5f8d
+size 198946
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..2943c6e2439906e4a20523da638b895ebc3555e4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a child's clock with a playful design. The clock has a white face with black numbers and hands, and it features a cartoon character in the center. The character is wearing a green hat and has large, round eyes. Surrounding the character are various food items, including a sandwich, a cupcake, and a slice of pizza. The clock is set on a wooden surface, and a hand is seen in the frame, reaching towards the clock. The style of the video is casual and child-friendly, with a focus on the whimsical design of the clock."
+    ],
+    "video_ids": [
+        "InGaVPreV7Q_28_0to190"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Child's clock, cartoon character, sandwich, cupcake, slice of pizza, hand? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a child's clock with a cartoon character (a pink squid-like figure) in the center. A hand places stickers of a sandwich, cupcake, and slice of pizza on the clock face at different hour positions. All specified objects are present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a clock's clock with a cartoon character ona styl figure-like figure) on the center. There hand is a of a sandwich, cupcake, and slice of pizza on the clock face. different times positions. The the objects are present and correctly the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..818309f223b7b1eb6c351d82522a11c9e0409408
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79d4cf55fc5c89901c7dfc8aa14610a5355c2f1b22d32fe519ed2446d2a87cde
+size 113177
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..9269c489416871fae518e456d5cda6cd76ef5a56
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news report about a school bus crash investigation. The style of the video is informative and serious. The video features a yellow school bus with the number 366 on the side, a stop sign on the back, and a black stripe running along the bottom. The bus is parked on a street with trees in the background. The news report includes a headline that reads \"DEVELOPING STORY DEADLY SCHOOL BUS CRASH INVESTIGATION DRIVER CHARGED WITH VEHICULAR HOMICIDE\". The video also includes a logo for the news network, which is GMA. The overall tone of the video is somber and investigative, as it reports on a tragic event and the subsequent investigation."
+    ],
+    "video_ids": [
+        "XXp4tRJKwUg_4_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Yellow school bus (number 366), stop sign, black stripe, trees, news network logo (GMA)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a yellow school bus with the number 366, a stop sign on its side, black stripes along its body, trees in the background, and the GMA news network logo at the bottom right corner. All elements mentioned in the description are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts ful a yellow school bus with the number 306, which stop sign, the side, a stripes on the body, and in the background, and a GMA news network logo. the bottom of corner. The elements match in the condition are present in match with the video content of}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..14984014d2aa3dd1ae6192f45a1604307c2aefec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e3583aa6287e83ff49eabb4e72d53b72569ea096a17cee24c545986d3198847
+size 147100
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..d377f71e782477da7c5f373a46bdd55b5a6d87db
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two young women are seated at a table in a library, engrossed in their work. The library is filled with books, and the women are surrounded by a quiet atmosphere. The woman on the left is wearing a black leather jacket and glasses, while the woman on the right is dressed in a yellow blouse. They are both focused on their notebooks, which are open in front of them. The scene captures a moment of concentration and academic diligence."
+    ],
+    "video_ids": [
+        "7lNYfZnp758_4_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two young women, a table, notebooks, books.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two young women sitting at a table with notebooks, which aligns with the 'Object(s)' condition. Although there are additional elements like a man walking by and bookshelves in the background, these do not contradict the core description and are consistent with a library setting. The presence of books on the shelves further supports the context.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two young women sitting at a table, notebooks and books ares with the 'Object(s)' condition. The there are additional books such books book in in in ashelves in the background, these do not contradict the core description and are acceptable with a library setting. The presence of books on the shelves further supports the library of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d73a38932021a67545381db3158762353878a00a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a4d82fdcebedeef2bf3e9657fbd4ea4bd8729a85c8c6b3858e19d65e8afdecac
+size 117064
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..f99d6885ffa243f2e1909b5e8ebd06917d44ab33
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a culinary journey featuring a delicious meal. The first frame shows a table set with a white plate of golden-brown fried chicken, a bowl of vibrant coleslaw, and a glass of milk. The second frame introduces a white bowl filled with creamy macaroni and cheese, accompanied by a fork. The third frame reveals a white plate of crispy fried chicken, a bowl of macaroni and cheese, and a glass of milk. The style of the video is a close-up, high-definition food photography that captures the textures and colors of the meal. The table setting is simple and elegant, with a white tablecloth and wooden placemats. The lighting is soft and warm, highlighting the appetizing appearance of the food. The video is likely intended to evoke a sense of comfort and indulgence, inviting viewers to enjoy a hearty and satisfying meal."
+    ],
+    "video_ids": [
+        "qyt1W_909ew_5_0to110"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: White plate of golden-brown fried chicken, bowl of vibrant coleslaw, glass of milk, white bowl of creamy macaroni and cheese, and a fork? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a white plate of golden-brown fried chicken, a bowl of vibrant coleslaw, a glass of milk, a white bowl of creamy macaroni and cheese, and a fork. All these items are visible and accurately represented in the scene without any contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showcasing showing a white plate with golden-brown fried chicken, a bowl of vibrant coleslaw, a glass of milk, a white bowl of creamy macaroni and cheese, and a fork. The these elements are present and match represented in the frames.\"\n any contradictions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6dc7a3b2d8204e7fc98c1121b5939fc482696b05
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1580545f190c2b2b29ca7488a443a6259802e20c430bf2fc13efd86492998648
+size 158318
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..f9bba43dd1d37f36238565cb37235d658332e5c4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a white Volkswagen SUV parked on a brick road. The car is positioned at an angle to the camera, allowing a clear view of its front and side. The license plate reads \"JV-037F\". The car is parked in front of a large building with a distinctive circular structure. The sky is overcast, suggesting a cloudy day. The overall style of the video is a straightforward, real-life depiction of the car in its environment."
+    ],
+    "video_ids": [
+        "HwyYCVXC4yc_18_0to202"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: White Volkswagen SUV, large building with a circular structure, license plate reading 'JV-037F'? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white Volkswagen SUV, which matches the description. In the background, there is a large building with a circular structure, and the license plate on the vehicle clearly reads 'JV-037-F'. All core elements of the description are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white Volkswagen SUV parked which matches the description. The the background, there is a large building with a circular structure, also the license plate of the car reads reads 'JV-037F'. The elements elements of the description are present in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..77bc1dc95c2f3caacf5ebe2a78a5059f42fce262
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b9369ed39296d201c277fb7ebec15c44f44cb54333b55170a345068a977aefea
+size 215393
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..edb2f92cf6290e8cc7910e6ba9336ac7eb17aa8d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a silver BMW i8 sports car in motion on a road. The car is sleek and futuristic in design, with a low and wide stance. The wheels are large and black, with a distinctive multi-spoke design. The car's body is adorned with blue accents, and the BMW logo is prominently displayed on the front grille. The car's headlights are on, and the car appears to be in motion, as suggested by the slight blur of the background. The road on which the car is driving is clear and well-maintained, with no other vehicles or pedestrians visible. The overall style of the video is dynamic and high-energy, emphasizing the car's speed and agility."
+    ],
+    "video_ids": [
+        "ucUMUWo_Now_15_0to117"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A silver BMW i8 sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a silver BMW i8 sports car, which matches the description. The car's distinctive design, including the blue accents, kidney grille, and sleek body, is clearly visible and consistent with the BMW i8 model. There are no elements that contradict the description of the object.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a silver car i8 sports car in which matches the description provided The car's sleek design, including its sleek accents and sleek grille, and overall body shape is consistent visible. consistent with the BMW i8 model. The are no additional in contradict the core.\"\n the object.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..444dbbbc8401648e180800425a49b41e8235abeb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ca2224a11c4fe4e4a2b3ba0a4646fc9d9b4f3fd2e98ec5c03f4d2e0c3ab4524e
+size 214766
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..fe3ee62fe8a7f88278fcaf5b876c6647911bf731
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man standing on a city street, taking a selfie. He is wearing a brown jacket and a white t-shirt. The street is lined with parked cars and buildings. The man is standing in front of a red stop sign. The sky is clear and blue. The man is smiling and looking directly at the camera. The video is shot in a casual, candid style."
+    ],
+    "video_ids": [
+        "JvyTcY52lTg_42_0to176"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a red stop sign, a brown jacket, and a white t-shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a brown jacket and a white t-shirt, standing on a street. A red stop sign is visible in the background. All core elements described are present in the video without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a brown jacket over a white t-shirt, standing on a street with There red stop sign is visible in the background. The the elements of in present in the video, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5c5afdcc4388b2d0af284f562aeb91dd7d4a3a15
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4778f7d1daf8725ce125f54a3f8e293b6a4aa27debe180c6e192b8e11c9b2e9
+size 191922
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..154b8c28357b797e5a809ae16755fb6dda477514
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young woman with long brown hair, wearing a light blue top. She is smiling and appears to be in a good mood. The background consists of white shutters, suggesting an indoor setting. The woman is gesturing with her hands, possibly explaining something or expressing enthusiasm. The overall style of the video is casual and friendly, with a focus on the woman's expressions and gestures."
+    ],
+    "video_ids": [
+        "vK1Pr8qp8Nc_2_0to128"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Young woman with long brown hair, wearing a light blue top.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young woman with long brown hair and a light blue top, which matches the description. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young woman with long brown hair, she light blue top, which matches the description provided The background is the elements in not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..07d9110e04dcd64dd1b0bb6d926d14802d4e95b8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1327-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a292fc770c51e78288571b971c5d53258b1cff3ecb60c9eea0535c864178ccc2
+size 168501
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..00c7ac0b17dd2f058fb87743f0d77b11c3575708
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the process of preparing a fresh salad in a large metal bowl. The first frame shows the salad ingredients being washed under a stream of water, with the vibrant colors of the vegetables contrasting against the metallic sheen of the bowl. The second frame shows the ingredients being chopped and mixed together, with the sharp blades of the knife glinting as they cut through the vegetables. The third frame shows the salad being seasoned with a drizzle of olive oil and a sprinkle of salt and pepper, the droplets of oil and seasoning adding a final touch to the colorful medley of vegetables. The style of the video is simple and straightforward, focusing on the preparation of the salad rather than any additional elements or distractions."
+    ],
+    "video_ids": [
+        "20ikVG4N_Q8_1_73to206"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Salad ingredients (vegetables), knife, olive oil, salt, pepper? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bowl of salad ingredients including chopped avocado, cucumber, red onion, cherry tomatoes, and parsley, which are vegetables. Olive oil is being poured over the salad, and black specks visible on the ingredients suggest salt and pepper. No knife is visible, but its absence does not contradict the condition since it's not required to be present. The core elements described are all present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bowl of chopped ingredients, chopped vegetables, bell, red bell, and tomatoes, and bell. which are typical. A oil is being poured over the salad, and a peppercks that in the vegetables suggest the and pepper. The other is shown in but the presence does not contradict the presence as the is not a for be shown in The video elements of are present present in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a8917726534bb03c89862dd4f384da3517e2d499
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:25480abf2ffc671f83437cd5417d08b3ef87685da013b488a8110e042f05bc6e
+size 238656
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..36ea833b81a88065756da960d325055db45ee1fe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a doll with long, colorful hair, wearing a dress with a floral pattern. The doll is standing on a wooden floor, and there is a blue chair and a pink wall in the background. The doll is looking directly at the camera, and the lighting in the room is soft and warm. The style of the video is realistic, with attention to detail in the doll's hair and clothing. The background is simple and uncluttered, allowing the doll to be the focal point of the video."
+    ],
+    "video_ids": [
+        "SL42Q6d2QGs_20_111to313"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A doll with long, colorful hair and a floral patterned dress.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a doll with long, colorful hair (pink, white, and yellow braids) and a floral patterned dress, which matches the description. The doll is the central focus, and no elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a doll with long, colorful hair,pink and yellow, and blue)ids) and a floral patterned dress. which matches the description provided The doll is the central focus, and the additional contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ad050633c6d65d3d5f5b54d7d331729e3c7c04a1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:028a1eb776da52745eb029c7c7a792aec4bcc865dd4d7958f0f7494f54964233
+size 189271
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..93186dde756ff763520b90c3ab50efaf1319f82c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a yellow off-road vehicle with large tires, driving through a field of tall grass and wildflowers. The vehicle is equipped with a roll bar and a spare tire mounted on the back. The scene is set against a backdrop of a partly cloudy sky. The vehicle appears to be in motion, as it navigates through the grassy terrain. The style of the video is realistic, capturing the vehicle's movement and the natural surroundings in detail. The focus is on the vehicle and its interaction with the environment, with no other people or objects in the frame. The video does not contain any text or additional graphics."
+    ],
+    "video_ids": [
+        "C5QRBazz6bY_60_0to151"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A yellow off-road vehicle with large tires, a roll bar, and a spare tire mounted on the back.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a yellow off-road vehicle with large tires, a roll bar, and a spare tire mounted on the back, which matches the description. The vehicle is clearly visible and these features are present throughout the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a yellow off-road vehicle with large tires, a roll bar, and a spare tire mounted on the back, which align the description provided The vehicle is parked depicted and the features are prominent in the frames.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1d923198d2caa07376b24ec2de4f634a39dd3777
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab30dd78e53d82867ab68e08425497f92f6a86ceb15b2ca82f1bd78bc8add28b
+size 332868
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..6da0afa9edc3d867fd1492d631476f071f6b11b0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a small dachshund dog sitting on a closed toilet lid. The dog is wearing an orange shirt and appears to be looking at a cell phone placed on the toilet lid. The dog seems to be interacting with the phone, possibly pressing buttons or swiping the screen. The setting is a bathroom, with a white toilet and a towel hanging on the wall. The dog's expression and posture suggest that it is engaged in a playful or curious activity. The style of the video is light-hearted and humorous, with a focus on the dog's amusing interaction with the phone."
+    ],
+    "video_ids": [
+        "pZ0Sez2LUiE_90_81to226"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small dachshund dog, an orange shirt, a cell phone? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a small dachshund dog wearing an orange shirt, sitting on a toilet with a cell phone placed in front of it. All core elements described in the 'Object(s)' condition are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a small dachshund dog wearing an orange shirt, sitting on a toilet. a cell phone placed in front of it. The the elements of in the questionObject(s)' condition are present and accurately depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1c3acb78b0f5acc105ac25ab10a8a347094aedc5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:638178c70a674ca0d36d5d7a6f308e1f3e7149527d4a12de40c9e4d8a0cf6f4a
+size 107777
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..be8cb74157f1cd02a690ad2ff24eebb39ae3433e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a vintage blue muscle car parked in front of a building with red and white walls. The car is a classic model, featuring a prominent hood scoop and a set of shiny chrome wheels. The car is parked on a paved surface, and the building behind it has a corrugated metal exterior. The car is the main focus of the video, and it is shown in three different angles, highlighting its design and details. The style of the video is straightforward and documentary-like, with no additional elements or effects. The car's position and the building's colors create a contrast that emphasizes the car's presence."
+    ],
+    "video_ids": [
+        "RKhQYKCxOwE_2_0to169"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A vintage blue muscle car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a vintage blue muscle car, specifically a Dodge Charger, with classic styling and features such as chrome trim, distinctive wheels, and a sleek body. The car is clearly the focal point and matches the description of a vintage blue muscle car.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a vintage blue muscle car parked which a classic Charger, which a features and a typical as a wheels and a headlights, and a hood body shape The car is parked the main point of matches the description of a vintage blue muscle car.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d96a2d77e1fe9c5dd3d1d4e8c32f4dfb4ff66c50
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:65d3b5a3975be911c7d07e7b471eb03e75646478dd434deb2cb2c3e3122dac9f
+size 103583
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..25bc7de84c59e9be5d98174b18b56679e7dc14bd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, three women are seen preparing and enjoying a variety of food and drinks. They are standing around a table filled with an assortment of items, including bowls, cups, and a cake. The women are engaged in conversation and seem to be having a good time. The table is set with a variety of food and drinks, including a cake, bowls of fruit, and cups of juice. The women are dressed casually and appear to be in a relaxed and comfortable setting. The video captures a moment of camaraderie and enjoyment among the women."
+    ],
+    "video_ids": [
+        "J28LX5f74dE_25_0to148"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three women, a table, bowls, cups, a cake, fruit, and juice.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three women at a table with various items including bowls, cups, a cake (visible on the screen behind them), fruit (bananas, strawberries), and juice (in pitchers and glasses). All elements mentioned in the condition are present, and there are no contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows two women standing a table with bowls items including bowls, cups, a cake,which in the table), the), fruit (appanas and oranges, and juice.in a). glasses). The the mentioned in the condition are present, and the are no contradictions.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8b7686fbe4cf5170177f5acfe2a9820af5601f72
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4047fb8d56e9105ea9c555cdba4d2a0e2d5deeb21b02fa51118e5745f1d70138
+size 222152
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..886f7888391cb8eac4869a548c45c201d2dc6b6f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man sitting in a black leather chair, wearing glasses and a brown leather jacket. He is holding his chin in his hand, suggesting a thoughtful or contemplative pose. Behind him, there is a green plant and a blue cup on a table. The background also includes a yellow sun with a red brain inside it, which is likely a logo or emblem. The style of the video appears to be a mix of live action and graphic design elements, possibly for an educational or informational purpose. The man's attire and the presence of the brain logo suggest that he may be a professional or expert in a field related to the brain or mental health."
+    ],
+    "video_ids": [
+        "EJbelcQxpQ4_4_0to131"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black leather chair, wearing glasses and a brown leather jacket, holding his chin. A green plant, a blue cup, a yellow sun with a red brain inside it.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man seated in a black leather chair, wearing glasses and a brown leather jacket, with his hand holding his chin. A green plant is visible to his left, a blue cup is on the table in front of him, and a yellow sun graphic with a red brain inside it is displayed in the upper left corner. All elements described are present and do not contradict the given description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man sitting in a black leather chair, wearing glasses and a brown leather jacket. with his hand on his chin. There green plant is visible in the right, and blue cup is on the table in front of him, and a yellow sun with with a red brain inside it is present in the background left corner of The elements match in present and match not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9aae8534dfcf72671a00d091be328c2fe49bd560
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b606a0afb7630d14e2ffc718b90eda2aea232363fa8b9befe138809d62211f8a
+size 171058
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..740c9986d993e0194eec18e36ce4feb9c5331580
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman and two children are seen sitting against a tree in a park. The woman is wearing a brown jacket and has red hair. The children are wearing casual clothing, with one of them holding a stuffed animal. The scene is set in a park with trees and grass, and the lighting suggests it's daytime. The style of the video is casual and candid, capturing a moment of family time in a natural setting."
+    ],
+    "video_ids": [
+        "GGoWZhDqlMs_28_45to173"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, two children, a tree, a stuffed animal.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman, two children, a tree, and a stuffed animal (a badger puppet held by one of the children). All core elements described in the condition are present and accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a woman and two children, a tree, and a stuffed animal.a plushger).). by the of the children). The the elements of in the condition are present, match depicted in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1f50996669fbdec3c7478e98a956ba29da2cdba0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1328-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2aed9e5c8008e5b179a021aa4b905fc15e72af39e15d84599f24e0faf643a75c
+size 145127
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..814c4156bc30b62a248d88ce44480da4f79402aa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are seen in a store. The man is wearing a blue shirt with a white logo on it, and the woman is wearing a red tank top. They are both smiling and appear to be enjoying their time together. The store has shelves filled with various items, and there is a sign that reads \"www.bn.com\". The man and woman are standing in front of a display that has a blue chair and a black bag on it. The woman is leaning on the man's shoulder, and they both seem to be looking at something off-camera. The overall atmosphere of the video is casual and friendly."
+    ],
+    "video_ids": [
+        "6amypbuookU_5_59to256"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a man and a woman as the main subjects, with the man seated and speaking while the woman leans beside him. Although there is another woman in the background, the core description of 'A man and a woman' is clearly fulfilled by the two individuals in the foreground, and no element contradicts this basic requirement.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a man and a woman standing the main subjects. standing the man holding and the, the woman stands in him, The there are a person in the background, the primary focus of theA man and a woman' is fulfilled met by the two individuals in the foreground.\"\n making the additional contradicts this.\"\n condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bc6f8ed3ddbd5000ce0068ab1eb3bfacaf302347
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e714381fdb08734ef1e11b6e5a3cc3336f7cc5fca5a815f9337ee78f587d7573
+size 195721
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..c6d4b15a7cb6aea1bd028cdc36a2f4b9b7a109ef
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a large, multi-layered burger being prepared on a wooden cutting board. The burger is stacked with several layers of beef patties, cheese, pickles, onions, and a tomato slice. The burger is placed on a bun, and the cutting board is covered with chopped onions. The style of the video is a slow-motion, high-definition food preparation video, focusing on the textures and colors of the ingredients. The lighting is bright and even, highlighting the details of the burger and the cutting board. The background is blurred, keeping the focus on the burger. The video is likely intended for a food-related channel or website, showcasing the preparation of a delicious and hearty meal."
+    ],
+    "video_ids": [
+        "1zRxcv6sALk_1_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large, multi-layered burger, beef patties, cheese, pickles, onions, tomato slice, bun? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a large, multi-layered burger with visible beef patties, cheese, pickles, onions, a tomato slice, and a bun. All the specified elements are present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a large, multi-layered burger with the beef patties, cheese, pickles, onions, and tomato slice, and a bun. The the elements elements are present, match depicted, the image,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..13aded3b1b1618622d5d4b1bfb7e6b7a7a21eb17
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17a71fad54fedfa85a9d903fdb43fb2f0ce6ab9ead486f216043f28c34dd5f66
+size 107911
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..5db07ae2cad21602e6d6457756374e1562c423d3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a delicious breakfast spread on a white plate, which is placed on a dining table. The plate is filled with a variety of breakfast items, including eggs, bacon, sausage, and toast. The eggs are cooked to perfection, with a runny yolk that adds to the overall appeal of the dish. The bacon is crispy and golden brown, while the sausage is well-cooked and flavorful. The toast is lightly toasted, providing a perfect base for the other items on the plate.  In addition to the main plate, there is also a small bowl filled with fresh fruit, including pineapple and watermelon. The fruit adds a refreshing touch to the meal, providing a contrast to the hearty breakfast items.  The overall style of the video is simple and straightforward, focusing on the food and the dining table. The camera angle is slightly elevated, providing a clear view of the plate and the food on it. The lighting is bright and even, highlighting the colors and textures of the food. The video does not contain any text or additional elements, allowing the viewer to fully focus on the breakfast spread."
+    ],
+    "video_ids": [
+        "e-DPnLIrl7w_1_71to210"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A plate with eggs, bacon, sausage, and toast; a small bowl with pineapple and watermelon.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a plate with eggs, bacon, sausage, and toast, as described. It also includes a small bowl with pineapple and watermelon (along with other fruits like grapes and honeydew). The presence of additional elements (like syrup and a cup) does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a plate with eggs, which, and, and what, which well. Additionally also includes a small bowl with what and watermelon,al with water fruits), mel), possiblydew mel The presence of additional fruits (fr the or other green) does not contradict the core description of}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ddb0cb4a14468833a1a5926de2513fa0ab1dfc85
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:903cd6ac183c8e0830f995f5557b52f29045a59f7955be7dce16fecb8a4a7397
+size 152237
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..ce99f2e4c66b2c87d72a3ec6cae5dd8f8e589ebb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a humorous take on a game show, featuring a man in a gray t-shirt with headphones, sitting at a table with a laptop and a stack of red cups. He is holding a red cup and appears to be in the middle of a drinking game. The laptop screen displays a game interface with various contestants' faces and their respective scores. The contestants are diverse, with different genders and ethnicities. The game interface also shows a timer and a scoreboard. The man seems to be enjoying the game, as he is smiling and laughing. The overall style of the video is light-hearted and entertaining, with a focus on the game and the contestants' reactions."
+    ],
+    "video_ids": [
+        "4kJmETiSB7w_12_238to399"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a gray t-shirt, headphones, laptop, stack of red cups, red cup in hand, game interface on the laptop, various contestants' faces, timer, scoreboard? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a gray t-shirt wearing headphones, sitting in front of a laptop with a game interface visible. He is holding a red cup, and there is a stack of red cups on the table. The game interface includes various contestants' faces and a scoreboard with scores. While a timer is not explicitly visible, the game interface and scoreboard are present, fulfilling the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man in a gray t-shirt wearing headphones, sitting at front of a laptop. a game interface on on The is holding a red cup, and there is a stack of red cups on the table. The game interface on various contestants' faces, a timer, a. The the timer is not explicitly mentioned, the presence interface suggests scoreboard suggest present, which the ' description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7aa76517c2e3f63a5b4e223a792cbe5f1a91407a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a2bcd6893a866657a3d0ab979cf5a49e0bd51a1b8c1279c2528785e54199af5
+size 184940
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..b92b6cf0f69e1f0d98a8adde032109a41b58157c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a white rectangular plate with three slices of chocolate chip banana bread. The bread has a golden brown crust and is speckled with dark chocolate chips. The plate is placed on a wooden table, and the background is blurred, focusing the viewer's attention on the bread. The style of the video is simple and straightforward, with a focus on the food item. The lighting is bright, highlighting the texture and color of the bread. The video does not contain any text or additional elements."
+    ],
+    "video_ids": [
+        "WsQ8xD9nF98_42_0to155"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white rectangular plate, three slices of chocolate chip banana bread.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a white rectangular plate with three slices of chocolate chip banana bread. The bread pieces are visibly topped with chocolate chips and arranged on the plate, matching the description. Additional elements like a hand and kitchen items in the background do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a white rectangular plate with three slices of chocolate chip banana bread. The description appears are topped topped with chocolate chips and banana neatly the plate, which the description provided There elements such the wooden or a setting are the background do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bd163f7b54924ab15d32c808fc0a9992347dc962
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:079dbfef98a1076ffff0b17c5081ffd741ace32b38e4a07525fcf52bb913da7c
+size 61026
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..8ae766bf3eb071d1aaa37b813af18a542b3a5722
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse of a water feature in a park or amusement area. The water feature is designed to resemble a natural river or stream, with red and brown rock formations and a large tree with a swing hanging from one of its branches. The water flows gently over the rocks, creating a serene and peaceful atmosphere. The time-lapse effect captures the movement of the water and the changing light conditions throughout the day. The overall style of the video is realistic and naturalistic, with a focus on the beauty of the water feature and the surrounding environment."
+    ],
+    "video_ids": [
+        "9we8JPMAH4M_91_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Water, rocks, a large tree with a swing hanging from it.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a body of water flowing through a channel, surrounded by red-colored rocks and artificial rock formations. A large, stylized tree with a swing (or hanging platform) is prominently featured on the left side of the frame. These elements match the core description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a flowing of water flowing over a landscape, surrounded by rocks rocks rocks. a-looking formations. There large tree stylized tree with green swing hangingwhich similar object) is visible featured in the left side of the frame. The elements align the description description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..20dff27f78fee57b3a92e806014eaba1d17bc4ca
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5b5eeff7ee1158e81b376efdf71a52cd179ce291c701a7d3851d338d3ae3977
+size 261432
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..c6e9ee689b08beb2604926b642d65a0b44ff28f8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a behind-the-scenes look at a car repair shop. In the first frame, a man in a blue and black jacket is standing at a workbench, holding a tool and examining it. He appears to be in the middle of a repair or maintenance task. In the second frame, another man in a black and orange striped shirt is standing next to him, observing the work. He seems to be interested in the process or offering advice. In the third frame, the man in the blue and black jacket is still at the workbench, but now he is holding a different tool. The workbench is cluttered with various tools and parts, indicating that they are in the middle of a complex repair or maintenance task. The background of the video shows a typical car repair shop, with red walls and shelves filled with tools and parts. The style of the video is candid and informal, capturing the everyday work of the mechanics in their natural environment."
+    ],
+    "video_ids": [
+        "36TELqLU0Nc_15_74to256"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men in jackets, a workbench, various tools and parts, a video background showing a car repair shop.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men standing at a workbench with various tools and parts, and the background clearly depicts a car repair shop setting with tool cabinets, license plates, and automotive-related decor. The men are wearing jackets, and the overall scene aligns with the described condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts two men in at a workbench with various tools and parts, which the background appears depicts a car repair shop,. toolsboxes and which plates, and other parts items. The men are wearing jackets, and the scene scene aligns with the description conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5420e3731f7e7770aac5f9baf024f353d97b66ca
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b3c5f098e484e21a7ebe6628b89d6dad462832073b9822cef3f30b0599023a6
+size 186176
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..5418c35f75491dec6fe6453430a5ebed97955372
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the elegance of a luxury car, focusing on its shiny silver hood and the distinctive emblem on the front. The emblem, a silver winged creature, stands out against the car's sleek design. The car is parked in a spacious garage, with a few people in the background, adding a sense of scale to the scene. The lighting in the garage highlights the car's polished surface, reflecting the emblem and creating a sense of depth. The overall style of the video is sleek and sophisticated, emphasizing the car's design and the emblem's intricate details."
+    ],
+    "video_ids": [
+        "-F1ppG0_iX0_3_16to202"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Luxury car with a shiny silver hood and a silver winged emblem? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a luxury car with a shiny silver hood and a silver winged emblem (the Rolls-Royce Spirit of Ecstasy), which matches the core description. The presence of background elements like people and interior lighting does not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close car with a shiny silver hood and a silver winged emblem,a emblem-Royce logo of Ecstasy). which matches the description description provided The car of the elements like the and lights details does not contradict the main and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a06b01264bf19f1ea7b423b7a6c04f536ef15df7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1329-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a25b0c1e11bd49a5a62b473a560e678ce60cd0ee8dbcb560145cad0114f70c08
+size 134812
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..265658a5649bce438d9249437ecff8f9f0ec5064
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person using a spoon to scoop a small amount of orange-colored substance from a glass bowl. The bowl is placed on a table, and the substance appears to be a type of sauce or condiment. The person is holding the spoon over the bowl, and the substance is being transferred from the bowl to the spoon. The table is covered with a white tablecloth, and there is a gold-colored decoration on the table. The decoration is made of gold foil and has a leaf-like shape. The person is wearing a white shirt, and the background is blurred, making it difficult to discern any additional details. The style of the video is simple and straightforward, focusing on the action of scooping the substance from the bowl."
+    ],
+    "video_ids": [
+        "K-4u3Avtee4_81_0to102"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a spoon, a glass bowl containing an orange-colored substance, a gold-colored decoration, and a white shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a person (hands visible), a spoon, a glass bowl containing orange-colored caviar, a gold-colored decorative element resembling a flower, and a white shirt worn by the person. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a person usingpart only), a spoon, a glass bowl containing an-colored substanceaviar, a gold-colored decoration element, leaves leaf, and a white shirt worn by the person. The elements elements of in present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c167e6ef7c6d74743bbc6a013a90dec3acbbf661
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5b2c5eac451cba01e35c744884987492794f3f14c1804cb4217539c3a4a049a
+size 166960
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..4216398d8655a22203dac4ad8aaf4af588c8951d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the dynamic process of a volcanic eruption. The first frame shows a large, snow-capped volcano with smoke billowing from its peak. The second frame shows the volcano erupting, with a large plume of smoke and ash rising into the sky. The third frame shows the aftermath of the eruption, with the volcano still smoking and the surrounding landscape covered in ash. The style of the video is realistic, capturing the natural beauty and power of the volcano. The camera angle is from a distance, allowing the viewer to see the full scale of the eruption. The colors in the video are predominantly blue, gray, and white, reflecting the natural colors of the volcano and its surroundings."
+    ],
+    "video_ids": [
+        "E64yyBaW2qo_36_28to168"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Volcano, smoke plume, ash cloud, landscape covered in ash? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a volcano with a visible smoke plume rising from its summit, and the surrounding landscape is covered in ash, matching the description. The visual elements are consistent with the specified conditions without any contradictory elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a volcano with a smoke smoke plume and from it peak, which the landscape landscape appears covered in a, which the ' of The presence elements of consistent with the core conditions.\"\n any contradictions features.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a479d75cbd96ee36ff40681f7cc8c07ac4561e80
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1365921e757af7994d73877cdda6537fef00781adcb09c9e3d6852d12da938ed
+size 177203
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..6456b973b9a6332ff96520d162473d8370207c0e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a bald man is seen sitting at a table, engaged in a conversation with another person. He is dressed in a blue jacket and a blue shirt, and his hands are clasped together on the table. The table is adorned with a vase of flowers, adding a touch of elegance to the setting. The man's expression suggests that he is listening intently to the other person, indicating a serious or important discussion. The overall style of the video is realistic, capturing a moment of interaction between two individuals in a formal setting."
+    ],
+    "video_ids": [
+        "X3Tb6IcYSIM_258_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bald man, another person, a table, a vase of flowers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a bald man seated at a table, engaged in conversation with another person (partially visible). A vase with flowers is also visible on the table in the background, fulfilling all the specified object conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a bald man sitting at a table with with in a. another person whopartially visible). There vase with flowers is also present on the table. front foreground. fulfilling all the conditions conditions conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..326d25650474f695678986d4bb725fee4c392c24
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:295e5df138f96c11807794febc4221ddd21e1c1cc4eeb291d65d47f81071205f
+size 80200
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..fb77f5ba3425d4dff8ce6010381b866da6dde915
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up view of the interior of a car, focusing on the steering wheel and dashboard. The steering wheel is black with a silver emblem in the center. The dashboard is equipped with a digital display showing various information such as the speed, fuel level, and temperature. The car's interior is well-lit, with the sunlight streaming in through the windows. The car appears to be a modern model, possibly a luxury vehicle. The style of the video is straightforward and informative, likely intended for a car review or advertisement."
+    ],
+    "video_ids": [
+        "PIRD-C-R1ro_1_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, digital dashboard display, car interior? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the steering wheel with the Hyundai logo, a digital dashboard display, and the overall car interior including seats and center console. All elements described in the condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a steering wheel, the logo logo, the digital dashboard display showing and parts interior car interior. the and part console. These elements are in the ' are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..17570b50cf522619fbad14ca52c27d8f8f7fa3db
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4fdff7ffb72db898552aaa248e2b87afd6f8cf4c9662c60dcb84319c7b5830e7
+size 206058
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..661724774e43d20c5fdebcb4b5ecf47e05556b48
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a car, focusing on the dashboard and steering wheel. The dashboard is black with red accents, and the speedometer and tachometer are prominently displayed. The speedometer shows a speed of 0 mph, while the tachometer indicates a speed of 1500 rpm. The fuel gauge is also visible, showing a full tank. The steering wheel is black and appears to be made of leather. The overall style of the video is a close-up shot, focusing on the details of the car's interior. The lighting is bright, highlighting the colors and textures of the dashboard and steering wheel. The video does not contain any text or additional objects."
+    ],
+    "video_ids": [
+        "u0Mqs_gq2og_32_24to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dashboard (black with red accents), speedometer (showing 0 mph), tachometer (indicating 1500 rpm), fuel gauge (full tank), steering wheel (black, leather-made).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a car dashboard with black background and red accents, a speedometer reading 0 mph, a tachometer indicating approximately 1500 rpm (around 1.5 on the scale), a fuel gauge at full (F), and a visible portion of a black steering wheel. All elements align with the description, with no contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dashboard dashboard with black and and red accents, a speedometer indicating 0 mph, a tachometer indicating  1500 rpm,though the1.5 on the scale), a fuel gauge showing full tank1), and a black black of a black leather wheel. The elements match with the description provided making no contradictions or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5c8906ec3b7f2c58bd2a5a1d9e14c53f41a71d98
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6dfc0b19ea329823e176f7f0529859490706e251bd25a77bca7c191a96e46d25
+size 206787
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..64177a457429f690c884dd38bf6e9e4242ac4312
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of triumph on a football field. A player, wearing a white jersey with the number 87, is in the foreground, his mouth open in a triumphant yell. He is surrounded by his teammates, also in white jerseys, who are running towards him in celebration. The crowd in the background is a blur of red and white, indicating a large turnout for the game. The atmosphere is one of excitement and victory."
+    ],
+    "video_ids": [
+        "X6Ec-hZaToA_12_30to164"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a white jersey with number 87, his teammates in white jerseys, and a crowd.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by prominently featuring a player in a white jersey with number 87, along with his teammates in white jerseys, and a crowd in the background. The presence of the NFL Draft 2021 logo does not contradict the core description, as it is an additional element that does not conflict with the main subjects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting featuring a player in a white jersey with the 87, his with his teammates in white jerseys. and a crowd in the background. The player of the crowd logo logo2023 logo in not contradict the core description and as it is a additional element that does not conflict with the main focus of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..38f479580616d86f5ffb2c5ef77916d315c0ec98
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f335032143890f0d28c1385d4465034f91e7e995d1e9583729c0e356248bba08
+size 268494
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..3a2060f88d4ab9ce58347f5ad035af14dc469cc2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a black car, focusing on the driver's side. The car is well-maintained and clean, with a sleek design. The driver's seat is black leather, and the steering wheel is also black. The dashboard is equipped with a touch screen display, and there are various buttons and controls on the steering wheel and the center console. The car appears to be parked in a garage or a driveway, as there is a green hedge visible through the window. The video is a straightforward, unembellished depiction of the car's interior, with no additional action or movement."
+    ],
+    "video_ids": [
+        "KO60sbcy9mg_8_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Driver's seat (black leather), steering wheel (black), touch screen display on the dashboard, buttons and controls on the steering wheel and center console? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a black leather driver's seat, a black steering wheel with visible buttons and controls, and a touch screen display on the dashboard. The center console also features various buttons and controls. All described elements are present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows shows a car leather driver's seat, a black steering wheel, a controls and controls, and a touch screen display on the dashboard. The center console also has buttons buttons and controls, The these elements are present in match depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6fdb95ac546b7bbde3d6b1d42375e20d2ed12868
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:73ff505611a6a80ee45ced951d4c31aff51c92674250f737ab860cab4d23a86e
+size 175113
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..ef2c9a83d224514d3b88c8764b52721ec797671e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person preparing a drink using a blender. The person is pouring a liquid, likely coffee extract, into a glass filled with ice. The blender is placed on a countertop, and there are other ingredients nearby, including milk and chocolate powder. The style of the video is instructional, with text overlaying the image to provide information about the ingredients used in the drink. The text reads \"INGREDIENTS 3 oz. 100% Pure Coffee Extract\". The overall setting suggests a home kitchen environment."
+    ],
+    "video_ids": [
+        "BH1QOGDR5Ms_8_0to122"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person, blender, glass filled with ice, liquid (coffee extract), milk, chocolate powder? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person pouring coffee extract into a glass filled with ice, with a blender (Vitamix) in the background. A glass of milk and a container of chocolate powder are also visible, matching the described ingredients. All core elements are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person pouring a extract into a glass filled with ice, which a blender andlikelyitamix) and the background. The glass of milk and a bowl of chocolate powder are also visible, align the objects objects. The the elements are present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c18be87b01212b6a0f428b2ffcb71a9c35cbdc4b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-133-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:35316f410527b263263b9d6a54ef576d3c7166b0dcf3b4e1a91dde0115c5489c
+size 156679
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..1a561fb72ddc77f19ed116c0d7c4ec8c3642a2ef
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a whimsical scene on a city street. A large, cartoonish statue of Bugs Bunny, dressed in a blue and white striped shirt and blue shorts, stands on the sidewalk. He is holding a carrot in his hand and is positioned next to a large, yellow surfboard. The surfboard is leaning against a building, which has a sign that reads \"Warner Bros Studio Store\". The street is lined with palm trees, adding a tropical vibe to the scene. The overall style of the video is playful and fun, with a focus on the Bugs Bunny statue and the surfboard."
+    ],
+    "video_ids": [
+        "aEe_epEaRNU_13_0to180"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bugs Bunny statue, surfboard, Warner Bros Studio Store sign? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a Bugs Bunny statue holding a surfboard in front of the Warner Bros Studio Store, with the store's sign prominently displayed above the entrance. These core elements are accurately represented, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two Bugs Bunny statue, a surfboard, front of a Warner Bros Studio Store sign which the store's sign visible displayed in. entrance. The elements elements match present represented, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ca97af96806418f25d27fcd02db73409087a0374
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3990cb171a202361216629e820d6da574573da428f96a24b25a1107a64563e61
+size 191999
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..d7e4c8c53ab19e88ce1bee614d2009eebdfa3337
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a thrilling moment in a football game. The main focus is a player from the Pittsburgh Steelers, who is in the process of catching a pass. He is wearing a black and yellow uniform, which stands out against the green field. The player is in the center of the frame, with the football securely held in his hands.  In the background, you can see other players from both teams, all in various states of motion. Some are running towards the action, while others are standing still, possibly strategizing their next move. The crowd in the stands is a blur of colors, indicating a large and enthusiastic audience.  The style of the video is dynamic and action-packed, capturing the intensity of the game. The camera angles and movements suggest a professional production, with multiple cameras capturing the action from different perspectives. The focus is on the player with the ball, but the background elements add depth and context to the scene. The overall impression is one of excitement and anticipation, as viewers wait to see what will happen next in this high-stakes game."
+    ],
+    "video_ids": [
+        "2IVMfxkZzXA_5_0to143"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Pittsburgh Steelers player, football, other players, crowd? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a Pittsburgh Steelers player (wearing jersey #84) holding a football, with other players (including a San Francisco 49ers player in jersey #35) and a crowd in the background. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts ful a player Steelers player runningdearing a number1)) running a football and running other players inin one player Francisco 49ers player) the #9)) and a crowd in the background. The elements elements of in present, contradiction.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..63ed3919f3d5f6abfbc1d9c01ca9ecee992edf03
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c62f60c66d985d23e76267dba977937cd59f0f51f6f54169ce5affe4bea0090
+size 282763
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..6a4074f3044ab427689f0a119f70f27b8b28d7df
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a maroon Honda Ridgeline truck parked on a gravel lot. The truck is positioned at a slight angle to the camera, allowing a clear view of its rear and side. The Honda logo is prominently displayed on the tailgate, and the model name \"Ridgeline\" is visible on the side of the truck. The truck is equipped with a tow hitch and a trailer brake controller, suggesting it is capable of towing a trailer. The license plate reads \"3421\" and the truck is registered in Iowa. The background features a brick wall and a clear sky, indicating the video was likely taken on a sunny day. The style of the video is straightforward and informative, focusing on the features of the truck without any additional context or embellishment."
+    ],
+    "video_ids": [
+        "2c-AKAvAth0_4_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A maroon Honda Ridgeline truck? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a maroon Honda Ridgeline truck from the rear view, with visible branding and model name on the tailgate, matching the description. The truck is stationary in an outdoor setting, and no conflicting elements are present that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a maroon truck Ridgeline truck parked a rear view. which the details and a name on the tailgate. which the description of The setting is parked, front outdoor setting, which there additional elements are present.\"\n would the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7685b7159c20d07b21d0d196686964161c8091d4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd9e75b397616a64dee855660826d1c21593aa20144764d995922ac852c9f47e
+size 181828
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..3d7c587a43707e9ef5480f11f4355a2f59fc6a91
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a grey hoodie standing in a kitchen. He has short brown hair and is looking directly at the camera with a slight smile. The kitchen is well-lit and features white cabinets, a stainless steel oven, and a microwave. The man appears to be in a good mood and is likely engaging in a casual conversation or recording a video. The style of the video is informal and personal, suggesting it might be a vlog or a casual social media post."
+    ],
+    "video_ids": [
+        "7x1G0U01rPk_53_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a grey hoodie with short brown hair, looking directly at the camera with a slight smile.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a grey hoodie with short brown hair, looking directly at the camera. Although his facial expression is more neutral or slightly serious rather than a clear smile, he is looking at the camera as described. The background and other elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a grey hoodie with a brown hair, looking directly at the camera with He the expression expression is not neutral than slightly serious rather than a smile smile, the is indeed directly the camera as described. The background includes setting elements in not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5371308f968a97c5a0d57a39c375aa2ce848d903
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b5c94f3726fe979cb3fd6651ccbf15d88ee166d5ab6373d6c87f9245922e280
+size 147349
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..c503cbc100c0f03b2ce6bab3d37849f42f4ee07b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two men are engaged in a conversation. The man on the left is wearing a black mask and a black baseball cap, while the man on the right has a white beard and is wearing glasses. They are standing in a room with a yellowish hue, and there are lights in the background. The man on the right appears to be speaking, as suggested by his facial expression and the direction of his gaze. The overall style of the video is casual and candid, capturing a moment of interaction between the two men."
+    ],
+    "video_ids": [
+        "HtkdNq-hiS0_10_79to297"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, one wearing a black mask and baseball cap, the other with a white beard and glasses.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men: one wearing a black mask and baseball cap, and the other with a white beard and glasses. The description matches the visual content accurately, with no contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two men. one wearing a black mask and a cap, and the other with a white beard and glasses. The presence matches the core elements of, with no additional or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f622c5c62b32af6d780509cd3c6f4280d99b7fc5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:430c9c54b475eaac3c0c8c4a51a59e8e807064043edb8bacc20e89f1fec37eab
+size 145194
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..d48695e9c2f311abfaf143a064cb1e668f68cc30
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a pot filled with a green leafy substance, possibly herbs or greens, being cooked. The pot is placed on a stove, and steam is rising from the pot, indicating that the contents are being heated. The style of the video is a simple, straightforward cooking tutorial, focusing on the pot and its contents without any additional context or background. The lighting is bright, highlighting the green color of the leaves and the metallic sheen of the pot. The video does not contain any text or additional elements, and the focus is solely on the cooking process."
+    ],
+    "video_ids": [
+        "lzhTfMesr2M_13_16to138"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A pot filled with green leafy substance (herbs or greens).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a pot filled with green leafy substances, which appear to be herbs or greens, submerged in boiling water. The description matches the core object in the video without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a pot filled with green leafy substance, which appear to be herbs or greens. as in water water. The leaves of the core elements in the video, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8bdcfa8fdd8da29c14bd9ec28956096af2d45f4a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85964285ab145ef364659ad8016966ff2b8569feb8e9a2ad117479089ea01b6e
+size 122145
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..d1c6af36c5fb2517fe5807ee849147488b65e168
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young girl with blonde hair and a blue dress is seen interacting with a giraffe in a zoo enclosure. The giraffe, with its brown and white spotted coat, is standing on a dirt ground and leaning over a wooden fence to reach the girl. The girl is holding out her hand towards the giraffe, seemingly offering it food. The scene takes place in a zoo enclosure with a stone wall in the background. The interaction between the girl and the giraffe is the main focus of the video, capturing a moment of connection between humans and animals."
+    ],
+    "video_ids": [
+        "cGE0S3V9uSc_36_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young girl with blonde hair and a blue dress, a giraffe with a brown and white spotted coat.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young girl with blonde hair and a blue dress (visible from 0:00 to 0:04) and a giraffe with a brown and white spotted coat (visible throughout the video). These core elements match the description. Additional elements, such as another child appearing later, do not contradict the primary description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a young girl with blonde hair and a blue dress interactingobject in the0:00 to 0:05) interacting a giraffe with a brown and white spotted coat (visible from the video). The elements elements match the description provided The elements, such as the girl and in, do not contradict the main description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..65ea7ee9989ee2ce57268e2a39c15d620b0b8bba
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d4e517d003e280e900c2536565bbea0a4f0df257aac10a0ebf0c5ba016d73454
+size 305867
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..85ceac130161c2018da761f9775cb070793cde1a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a tie-dye shirt standing in a living room. He is looking directly at the camera with a slight smile on his face. The living room is well-lit and has a modern design, with a large window allowing natural light to fill the space. The man is the main subject of the video, and his tie-dye shirt adds a touch of color to the scene. The overall style of the video is casual and relaxed, with the man appearing comfortable and at ease in his surroundings."
+    ],
+    "video_ids": [
+        "HWw9Hdq8qeE_1_89to227"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a tie-dye shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man wearing a tie-dye shirt, which matches the core description. The background and other elements do not contradict this primary object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person wearing a shirt-dye shirt, which matches the description description. The shirt and additional elements in not contradict the description condition.\"\n.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2ba76c2a2f1667cf2cbfbd0063261f93b422f120
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1330-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff69a4ed38371a58a39897e6964ec73d7deb064be195ba8c8740a7527248f7d7
+size 137432
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..a899146105f2f5739a3eb46adca88bb5721cff23
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a drone shot of a river flowing through a dry, arid landscape. The river is surrounded by brown, dry grass and shrubs, with a few scattered trees. The river itself is a deep blue, contrasting with the brown surroundings. The drone flies over the river, providing a bird's eye view of the landscape. The video is shot in daylight, with clear skies. The style of the video is realistic, capturing the natural beauty of the landscape."
+    ],
+    "video_ids": [
+        "2X-dWa2MhxE_14_56to227"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: River, brown dry grass, shrubs, scattered trees? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a winding river flowing through a landscape dominated by brown dry grass, with scattered shrubs and a few trees visible along the riverbanks and in the distance. These elements align with the described 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a river river with through a landscape that by brown dry grass, shr scattered shrubs and trees few trees.. the riverbanks. in the distance. The elements align with the description 'Object(s)' condition, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5ce96bff01ab187d513914f52e14927843a15838
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:932f218ce6beb2873b0a318be5e9bb9ca0057aa056390d401f4d1c180384433f
+size 255440
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..0e436f8b625a5ab0aa190eced25c66d553f4326e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a child are seen riding small motorcycles on a driveway. The man is on the right, wearing a black t-shirt and a white helmet, while the child is on the left, wearing a blue helmet. They are both riding small motorcycles, with the man on a pink one and the child on a blue one. In the background, there are other motorcycles parked, and a barbecue grill can be seen. The scene is set in a suburban area, with a fence and trees in the background. The video captures a fun and playful moment between the man and the child, as they enjoy their ride on the small motorcycles."
+    ],
+    "video_ids": [
+        "cGu7FMFHlks_4_1002to1218"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a child riding small motorcycles.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a child both seated on small motorcycles, wearing helmets, and positioned as if ready to ride. The core description of 'A man and a child riding small motorcycles' is accurately represented, even though the motorcycles are stationary in the video. Additional background elements (like other bikes, a grill, and a tree) do not contradict the primary condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a child riding riding on small motorcycles, which helmets, which riding as if they to ride. The motorcycles elements of aA man and a child riding small motorcycles' is fulfilled represented in with though the motorcycles appear not in the image. The elements elements likelike the parked and a fence, and a fence) do not contradict the main focus.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6756e4ca25e281870450d96a369e0b405be54487
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:acc0cead915d48e51fa1ee24f148f0d9cdc5126aa8e3e2b7421371dcac0a0ec9
+size 199076
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..55da6f64df728e2064a0e8fc6c8ae393b61837a6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene moment in nature, featuring a white dog with striking blue eyes. The dog is seen in three different positions, each time looking directly at the camera with a calm and curious expression. The first frame shows the dog sitting on the ground, surrounded by lush green foliage. In the second frame, the dog has moved slightly to the left, still sitting amidst the greenery. The third frame captures the dog in a standing position, still surrounded by the same verdant environment. The backdrop of the video is a tranquil lake, adding to the peaceful ambiance of the scene. The dog's gaze and the natural setting create a sense of connection between the viewer and the animal, evoking a feeling of tranquility and harmony with nature."
+    ],
+    "video_ids": [
+        "R3p8po2HO8k_72_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white dog with striking blue eyes? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a white dog with striking blue eyes, matching the description. The dog is the central focus, and its features are consistent with the specified criteria.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video features depicts a white dog with striking blue eyes, which the description provided The dog is positioned central focus in and its blue, consistent with the given characteristics.}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bcb08338cef15fbe957c69532a0b084a8acc4ab0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c5f7fef97467d6f13dab95fcb9c8e056c5a64160f2946c4324f3b0eb2d014c25
+size 186223
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..39c22043ee695222bf1385ba32c9b223861cefc4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two young boys are seen decorating a colorful, multi-layered birthday cake. The cake is placed on a dining table, and the boys are using colorful icing to add decorations. The cake has a vibrant mix of colors, including red, yellow, and blue layers. The boys are focused on their task, carefully applying the icing to the cake. The scene is set in a room with a white wall in the background. The video captures the joy and creativity of the boys as they celebrate the special occasion."
+    ],
+    "video_ids": [
+        "1K4V7OSDytI_19_645to778"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two young boys? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two young boys standing at a table, interacting with a colorful cake and candles. Their presence and actions align with the 'Object(s)' condition of the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two young boys interacting side a table, engaging with a colorful,. using. The presence and actions align with the descriptionObject(s)' condition described the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c319c5b6401526f15c9294ba3cc33d2cbe836be6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4bfc723ede81f60e6f17d129973338254e7115c3d0aa673cf4c592bb945ae235
+size 172809
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..05aed83b21a00d5eefe91352b56b0307d4fbc592
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures an intense moment in a football game. A player in a blue uniform is leaping into the air, reaching out to catch a football. His body is stretched out, and his arms are fully extended towards the ball. The football is in mid-air, just within his reach. Another player, wearing a white uniform, is running towards him, trying to block the catch. The player in blue is wearing a helmet, and his focus is solely on the ball. The background shows a crowd of spectators, their faces a blur of anticipation. The style of the video is dynamic and action-packed, capturing the excitement and tension of the game."
+    ],
+    "video_ids": [
+        "00mbT1YOJME_0_0to168"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a blue uniform, a player in a white uniform, and a football in mid-air.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a blue uniform (New York Giants, #21) catching a football in mid-air, while another player in a white uniform (Dallas Cowboys, #39) is attempting to defend. The football is clearly visible in mid-air, being secured by the player in blue. The scene matches the core description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a player in a blue uniform,number York Giants) number13) and a football, mid-air, while another player in a white uniform (op Cowboys, #2)) is nearby to block. The football is clearly visible in mid-air, and caught by the player in blue. The presence is the description description of any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..70d5cf51bd5d64c44d283cd13bde58593b7971a4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04a9f63b99ac55d2762d7a6309e8ba8b0a174780335315ed8566ecd85e74a5ae
+size 274544
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..e7042515875d94332ff54b1bacab7abde96b89d1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The scene is set on a football field, with players from two teams, the New England Patriots and the San Diego Chargers, in action. The quarterback of the Patriots, wearing a blue jersey with the number 12, is in the process of calling a play, pointing towards the right side of the field. His teammates, also in blue jerseys, are lined up in a formation, ready to execute the play. On the other side, the Chargers, dressed in white jerseys, are preparing to defend. The crowd in the background adds to the atmosphere of the game. The video is a snapshot of the intensity and strategy involved in a football match."
+    ],
+    "video_ids": [
+        "bRE3YIN-07k_34_298to475"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Quarterback of the New England Patriots, teammates of the Patriots in blue jerseys, and the San Diego Chargers in white jerseys.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a quarterback wearing a New England Patriots jersey (number 12), surrounded by teammates in blue jerseys and opposing players in white jerseys (San Diego Chargers). The scene is consistent with the described condition, depicting a football game with the correct team uniforms and positioning.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows football scene and a blue England Patriots jersey,blue 12) along by teammates in blue jerseys, players players in white jerseys.likely Diego Chargers). The scene is set with the description scenario, as a football game scenario the correct teams uniforms and players.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..aedf75e0ef537a2a6e6bc51d74766c2d9c45a6e1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c62c1caf0224fda50cd5d0b96af15b28f3e0688f5300f9681ed264adac968319
+size 193042
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..d9ce149e9d2b557288d4f19dd4728d3746670044
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen in a citrus orchard, surrounded by orange trees. The person is holding a bucket filled with a green substance, which appears to be a type of fertilizer or pesticide. They are using this substance to care for the orange trees, likely to help them grow and produce fruit. The orchard is lush and green, with many orange trees bearing fruit. The person is dressed in casual clothing suitable for outdoor work. The overall style of the video is a real-life, documentary-style depiction of agricultural practices."
+    ],
+    "video_ids": [
+        "VphA42mIibI_7_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a bucket filled with a green substance, orange trees? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person holding a white bucket filled with a green granular substance, positioned among orange trees laden with fruit. The scene matches the described elements without contradiction, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a person holding a bucket bucket filled with a green substanceular substance. which in orange trees. with ripe. The presence align the description elements of any.\"\n fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d695d03c0d4a2d03809473ce6ca1492fe9415a06
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6e25f5d3b569c2d21e5bb82e8840bcd51e8b1f900a289b29f2f777ea552671ea
+size 263621
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..8f2f379f4752918d8bf5c7f44cbd7ccfe64cdb65
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a plate of food, which includes a bowl of white sauce, lemon wedges, and a garnish of green herbs. The food appears to be a type of fried fish sticks or breaded meat sticks. The style of the video is a simple, straightforward food presentation, with a focus on the textures and colors of the food and garnishes. The lighting is bright and even, highlighting the details of the food and the plate. The video does not contain any text or additional elements, and the focus is solely on the food and its presentation."
+    ],
+    "video_ids": [
+        "m1DuOo5LS1o_2_0to151"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bowl of white sauce, lemon wedges, green herbs, fried fish sticks or breaded meat sticks.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bowl of white sauce (likely yogurt or tzatziki), lemon wedges, green herbs (parsley), and fried breaded items that resemble meat sticks or fish sticks. All elements described in the condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a plate of white sauce,likely a or aatziki), lemon wedges, green herbs,possiblyley or and what fished meat that resemble fish sticks or fish sticks. The these match in the condition are present in match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a0e8831a42f98270d6c0ef2279a0aabe495942ac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1331-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6276ef0ece86be166384048ff17d890f29033bfbe1267f8c4e5f8600b4f1876d
+size 65795
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..3cf8e260ef759d05fd6c06179d277b1bb7dca300
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man riding a mountain bike on a trail through a forest. He is wearing a black and gray camouflage shirt, black pants, and a white helmet. The man is gesturing with his hands as he rides, possibly explaining something or giving directions. The trail is surrounded by trees and the ground is covered with leaves and dirt. The man appears to be in motion, suggesting that the video is capturing a moment of his ride. The style of the video is realistic and it seems to be shot in natural light, which gives the scene a natural and authentic feel."
+    ],
+    "video_ids": [
+        "9uw3G20N91I_38_21to143"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man on a mountain bike, wearing a black and gray camouflage shirt, black pants, and a white helmet.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The man in the video is on a mountain bike, wearing a black and gray camouflage shirt, black pants, and a white helmet. These elements match the description provided. The video does not contradict the core description, even though there are additional elements like the forest background and other people in the distance, which are acceptable as per the guidelines.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video in the video is riding a mountain bike, wearing a black and gray camouflage shirt, black pants, and a white helmet. The elements match the description provided. The setting shows not contradict the description description, and though the are additional elements like the greened and the details in the distance, which are not as long the instruction.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cb96ada10949a2cfcf7e27a7791b4dbee10b07fb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a3f273ed27232df721a8a9f302148b22de7b4734969d6eb0bc0b652809492a5f
+size 381583
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..78c8c3cf546f20c20791420034bd815bcc28961f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a hockey game. Two players, one in an orange and blue uniform and the other in a white and black uniform, are engaged in a fierce competition on the ice. The player in the orange and blue uniform is in the process of taking a shot at the goal, his body leaning forward with intensity. The player in the white and black uniform is attempting to block the shot, his body positioned defensively. The scene is set against the backdrop of a hockey rink, with the goal and the spectators visible in the distance. The video is a freeze-frame of a high-energy moment, capturing the tension and excitement of the game."
+    ],
+    "video_ids": [
+        "3kgLg3htjuo_6_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two players (one in orange and blue, one in white and black), a goal, spectators? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two hockey players: one in orange and blue (Edmonton Oilers) and one in white and black (Anaheim Ducks), as described. Spectators are visible in the background, and while a goal is not directly visible, the context of a hockey game implies its presence. The core description is accurately fulfilled.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts two players players, one in an and blue,numbermonton Oilers) and one in white and black (likelyheim Ducks). which well. Theators are not in the background, and the there goal is not directly shown, the setting of the hockey game implies the presence. The players elements is largely represented.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..aabf39545effe83e84b1766c73f2a55ce0491cb6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41ba69d187321bec5fded96d41127ca7d5f0992042b3d3f97a05b652a700e8b7
+size 238204
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..e7d115ddb097f9a60aaa4b9298500a8a1cd16910
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a koi fish swimming gracefully through a pond adorned with floating lily pads. The water is clear, allowing a view of the aquatic plants beneath the surface. The koi, with its distinctive white body marked by black spots, moves smoothly from left to right across the frame. As it swims, the fish's fins gently undulate, creating subtle ripples on the water's surface. The lily pads, varying in shades of green and some tinged with autumnal hues, float leisurely atop the water, adding a tranquil element to the setting. The overall atmosphere is peaceful, with the gentle movement of the fish and the stillness of the lily pads contributing to a calming visual experience."
+    ],
+    "video_ids": [
+        "d2727f1e3cfc83820909d43b218ab3731ff303f56dfcbf770cd60c39dc7e3c12"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Koi fish, lily pads? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a koi fish swimming in a pond with lily pads floating on the surface. The fish is prominently featured, and the lily pads are visible throughout the frames, matching the specified objects in the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a Koi fish swimming among a pond surrounded lily pads. on the surface. The presence and prominently featured in and the lily pads are visible, the scene, fulfilling the description objects. the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e8b8c060b3297097a75bdf86dc37b12d48d098f6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89e725974bdd7dc94eb9ab7dc506b4cdfce608ebdc040f089c239a331de650ee
+size 203647
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..bcc627d5a36e97fea94623dfd0a55e490fce803c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of triumph for a football player. The player, wearing a blue jersey with the number 10, is seen waving to the crowd with a broad smile on his face. He is standing on a football field, which is visible in the background. The crowd, although not directly visible, can be inferred from the player's gesture. The player's joyous expression and the waving gesture suggest a successful game or a personal achievement. The overall style of the video is dynamic and captures the essence of the sport and the spirit of the player."
+    ],
+    "video_ids": [
+        "JSRhQwn2V2U_75_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A football player wearing a blue jersey with the number 10, a crowd (inferred from the player's gesture).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a blue jersey with the number 10, which matches the description. The player is waving, implying interaction with a crowd, which is inferred as part of the condition. The background elements, such as the stadium lights and screens, are consistent with a game setting where a crowd would be present. No conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a blue jersey with the number 10, which matches the description. The player is also, which the with a crowd, which is inferred from part of the description. The background suggests, such as the field and and the, suggest consistent with a football setting, a crowd would be present. Therefore additional elements are observed.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6c160f2bd7f2a7018ab14b965ea19e8a86280d4d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bbdb43f0ab30900e7a73a8e7c199165b761e371857b3b5e3f1f1543b7658ae05
+size 190698
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..192c4d562f3d896468d93ca8378c600e0f0f3f66
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man standing in a grassy field with trees in the background. He is wearing a gray t-shirt and has a beard. In the first frame, he is looking directly at the camera with his arm down. In the second frame, he is speaking and his arm is slightly raised. In the third frame, he has his arm fully extended towards the camera, gesturing as he speaks. The style of the video is casual and informal, with the man appearing to be engaged in a conversation or presentation. The setting is outdoors and the lighting is natural, suggesting that the video was taken during the day."
+    ],
+    "video_ids": [
+        "1jB16_IKOGY_38_676to866"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man as the central subject, speaking to the camera and gesturing with his hand. The description of 'A man' is accurately fulfilled, as there are no contradictions or misrepresentations of the main object in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a man standing the main figure. standing and the camera in makinguring with his hands. The background of theA man' is fulfilled represented by as the is no additional or additionalrepresentations in the subject subject in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5f076acd817090e8fb96b0b0e6b3185e1961e4b4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc4473e9bf2e9c7df94229a76a157a9e6f4ada3bfcc40ba72d39191047848ae4
+size 169372
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..62e6acf1d47f650c27eb7ae7d26ce32ed9600ea6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two men are seen in a playful and humorous scene. The first man, dressed in an orange jacket, sports a white beard and a red Santa hat, giving him a festive appearance. He is holding a pink sports ball, perhaps preparing for a game or a playful throw. The second man, wearing a black jacket, has a unique accessory - a mask of a famous politician's face. He is gesturing with his hands, possibly in the middle of a conversation or an explanation. The background reveals a grassy field, suggesting an outdoor setting, and a fence can be seen in the distance. The overall style of the video is light-hearted and fun, capturing a moment of camaraderie and amusement between the two men."
+    ],
+    "video_ids": [
+        "2quvvKYb22Y_9_169to349"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, one in an orange jacket with a white beard and red Santa hat, holding a pink sports ball; another in a black jacket with a politician's face mask.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men as described: one wearing an orange jacket, white beard, and red Santa hat, and the other wearing a black jacket with a politician's face mask. Although the pink sports ball is not clearly visible or held by the man in the Santa outfit in the frames provided, the core description of the two individuals and their attire is accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two individuals, described in one wearing an orange jacket with a beard, and red Santa hat, holding the other in a black jacket with a politician's face mask. The the second object ball is not clearly visible, held by the man in the Santa outfit, the provided provided, the description elements of the two men and their attire is accurately represented.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..713543b4395c179f59ee1f156dd965fcac23b06e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d51de8506eddf0b8c744a2692268dc2d29349328aee960a2187968afeb0d33eb
+size 170473
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..071536b02d98e3da070df6585a549351ccaf30cb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person preparing a meal in a kitchen. The person is using a silver tongs to pick up a serving of spaghetti from a white plate. The spaghetti is yellow and appears to be cooked. The person is holding the plate in their left hand and the tongs in their right hand. The kitchen counter is visible in the background and there are other dishes on the counter. The style of the video is a close-up shot of the person preparing the meal, focusing on the action of picking up the spaghetti. The lighting in the kitchen is bright and the colors are vibrant."
+    ],
+    "video_ids": [
+        "PXNka1s4lp0_26_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person, silver tongs, white plate, yellow spaghetti? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person using silver tongs to serve yellow spaghetti onto a white plate, which matches the core description. Additional elements like a watch, rings, and background kitchen items do not contradict the specified objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person using silver tongs to lift yellow spaghetti onto a white plate. which matches the description description of The elements such the bowl on a, and a details items do not contradict the main objects.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b02da89ccb9670ebaef0503756f5ca18874fbfb3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a67dbcf123660f8d3814d3190af3318529a009e1095243fb18b05e097048e04
+size 162382
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..eb8ae5b94cb1bf726e1d355ce54ca73a4abe04fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a BMW car, focusing on the gear shift and the center console. The gear shift, which is the main subject of the video, is a sleek, black and silver design with the BMW logo prominently displayed. It is positioned in the center console, which is also black and silver, and features a variety of buttons and controls. The gear shift is shown in three different positions: neutral, drive, and reverse. The video is shot from a high angle, providing a clear view of the gear shift and the center console. The style of the video is straightforward and informative, providing a detailed look at the gear shift and its functions."
+    ],
+    "video_ids": [
+        "Ga_Uzaqj3qc_34_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Gear shift, center console? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the gear shift lever and the surrounding center console area of a car, including buttons and controls. The gear shift is prominently displayed with the BMW M logo, and the center console features various controls and a rotary dial, matching the described 'Object(s)': Gear shift, center console.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a gear shift and and the center center console, of a car. which the and controls. The presence shift is prominently displayed in the BMW logo logo, and the center console includes various buttons and buttons gear dial, which the description elementsObject(s)' Gear shift and center console.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..92353bba97cdfa1e1239f39ca6c09bfa2955eecd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1332-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:848e9cfcfb776fd6fa96fa79f800dfad794142b0bf7a6bcd3fb2f52e27d8136b
+size 130340
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..273b4dc0832a812869fc48c73a663f3507507902
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a man standing in front of the White House. The man is dressed in a suit and tie, and he appears to be speaking to the camera. The background shows the White House, which is a large white building with columns. The man is identified as Peter Navarro, Director of the National Trade Council. The video is likely discussing a political or economic issue related to the White House or the National Trade Council. The style of the video is a standard news segment, with the man standing in front of the camera and providing information. The focus is on the man and his message, with the White House serving as a backdrop to the story."
+    ],
+    "video_ids": [
+        "56-Qt19I2Wg_0_554to738"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man dressed in a suit and tie, identified as Peter Navarro, Director of the National Trade Council.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed in a suit and tie, and the on-screen text clearly identifies him as Peter Navarro, Director of the National Trade Council. The visual and textual elements align with the description, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in in a suit and tie, which the text-screen text identifies identifies him as Peter Navarro, Director of the National Trade Council. The setting and textual elements in with the description provided fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ad60b5eebef5249bb006af133b2969b3920b72a3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a719edef04828f0c9f456513648a3bae92660e8566d3d49423e700e813113248
+size 186496
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..9daa063d31a6941029c3625d8910d4b3143ef9bb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a large orange tiger is seen running towards a small gray rabbit. The tiger is in motion, with its mouth open as if it's about to pounce. The rabbit, on the other hand, is standing still, looking up at the tiger with a surprised expression. The scene takes place in a lush green forest, with trees and bushes surrounding them. The tiger is running on a dirt path, while the rabbit is standing on the grass. The overall style of the video is animated, with vibrant colors and detailed backgrounds. The animation style is realistic, with attention to the textures and movements of the characters. The video captures a moment of suspense and action, as the viewer waits to see what will happen between the tiger and the rabbit."
+    ],
+    "video_ids": [
+        "85jJ_Rt30SY_58_24to216"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large orange tiger and a small gray rabbit? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly features a large orange tiger and a small gray rabbit, which matches the 'Object(s)' condition. The tiger is seen emerging from a house and interacting with the rabbit, who is holding a hockey stick. There are no elements in the video that contradict the presence of these two main characters.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a large orange tiger and a small gray rabbit, which are the 'Object(s)' condition. The tiger is depicted in from behind tree, walking with the rabbit, which is sitting a can puck. The are no additional in the video that contradict the description of these two animals objects.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d49d5d537a036fa73c38860c74884517b07e69dd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f045b71fd6218776da2bf0407f0d82978818767c4993b04f6b4aaae1b07bd77
+size 294013
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..c940b1ae02aa4bdd616f30fb5b8b1afa1da8b9db
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene featuring a cat sitting on a wooden bench by a pond. The cat, wearing a pink collar, is intently watching two koi fish swimming gracefully in the water. The koi fish, one with a vibrant orange and white pattern, moves fluidly across the surface of the pond, creating gentle ripples. The setting appears to be an indoor or semi-outdoor space, with wooden structures and a calm atmosphere. The cat remains stationary, its attention fixed on the fish, occasionally twitching its tail slightly. The overall mood of the video is peaceful and contemplative, highlighting the quiet interaction between the cat and the koi fish."
+    ],
+    "video_ids": [
+        "24350b4ed3030262ef44a88f80c3986be14bdd6b54ba2bf0849a45572637eb39"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Cat, wooden bench, koi fish, pond? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a cat sitting on a wooden bench, observing koi fish swimming in a pond. All core objects mentioned \u2014 cat, wooden bench, koi fish, and pond \u2014 are clearly visible and consistent with the description. There are no contradictions or misleading elements that detract from this core depiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a cat wearing on a wooden bench, observing koi fish swimming in a pond. The the elements ( in the, wooden bench, koi fish, and pond \u2014 are present visible and present with the description. The are no additional or additional elements.\"\n wouldact from the core scene.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0a7c7a44f5e03ea68d753ac9c2d4466cce658e41
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd9a7c1d18f5bbbf141784e3e360e884f17b8f075f5670f47c442be69e7ef61f
+size 229011
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..1dd9d55ecb417c06096fc9318a14e710da201727
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up, high-resolution, and colorful culinary presentation of a dish being prepared. The dish appears to be a type of Asian cuisine, possibly a stir-fry or a grilled meat dish, garnished with green onions. The dish is presented on a white plate with a floral pattern, and the plate is placed on a dining table. The table setting includes chopsticks and a small bowl of sauce, suggesting that the dish is meant to be eaten with chopsticks and accompanied by the sauce. The style of the video is that of a food blog or a cooking tutorial, with a focus on the presentation and the details of the dish. The lighting is bright and even, highlighting the textures and colors of the food. The video does not contain any text or narration."
+    ],
+    "video_ids": [
+        "VUknQygFYUg_10_0to158"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A close-up of an Asian cuisine dish (stir-fry or grilled meat), green onions, chopsticks, and a small bowl of sauce.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of an Asian cuisine dish, which appears to be stir-fried shrimp, garnished with green onions. Chopsticks are placed beside the plate, and there is a small bowl of sauce visible in the background. These elements align with the specified 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a Asian cuisine dish, which appears to be a-fried, with garnished with green onions. Theresticks and visible on the plate, and there is a small bowl of sauce in in the background. The elements align with the description conditionsObject(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cb1f63e8d410d06b8c78df2d1fb4a92eca49bdb8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e99bf56c750498dc05077bc97952efae6a81e641e3ac180738e72a3a4ec32bfb
+size 185722
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..4c3569e7e7d830a525fa186b1db2eb502ecceee0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a stylized, animated sequence featuring a robot character in a futuristic setting. The robot, with its sleek, black and gold design, is the central figure in the video. It is shown in three different frames, each capturing a different action. In the first frame, the robot is seen walking confidently down a neon-lit street, its metallic body gleaming under the artificial lights. The street is lined with towering buildings, their windows reflecting the vibrant colors of the neon signs. In the second frame, the robot is seen leaping into the air, its body arched gracefully as it soars above the cityscape. The third frame shows the robot landing on a rooftop, its feet firmly planted on the concrete surface. The video is a dynamic display of motion and technology, set against the backdrop of a futuristic city."
+    ],
+    "video_ids": [
+        "IgscAOoBtAc_7_21to144"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A robot character with a sleek, black and gold design.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a robot character with a sleek, black and gold design, which matches the description. The robot's form is stylized with metallic sheen and glowing edges, consistent with the black and gold aesthetic. While there are other robotic figures and background elements, the main character clearly fulfills the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a robot character with a sleek, black and gold design, which matches the description provided The character is appearance is consistentized and a elementsen and a elements, consistent with a ' and gold color. The the are additional elements elements in elements elements, the primary focus clearly adherfills the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f9987adc22b593dd4d1c74f9b4da10370cc5978e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:239a0c322bc64e46afc91acde2abba8204f35a6e65a93f8b9042fb3c471f7918
+size 305360
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..9a87f4c8a60e65c33bc6213f686f15c1dbc7978a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen in a colorful and vibrant setting. He is wearing a white hat and a plaid jacket, and he has a surprised or shocked expression on his face. The background features a blue mannequin and a pink sign with the word \"FLOW\" written on it. The overall style of the video is lively and energetic, with a focus on the man's reaction to something in the scene. The colors are bright and bold, creating a visually striking image. The man's surprised expression adds a sense of drama and excitement to the video."
+    ],
+    "video_ids": [
+        "Bpy4qJzqjNI_6_0to169"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a blue mannequin, a pink sign with 'FLOW' written on it.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a white hat and plaid jacket, a blue mannequin in the background wearing a white bra, and a pink sign with 'FLOW' written on it. These elements match the description provided, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video features shows a man wearing a white hat and aaid jacket, which blue balloonnequin, the foreground, a white hat, and a pink sign with 'FLOW' written on it. The elements match the description provided, fulfilling the additional elements are present.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e291c3776b1ec2779876ec52696c642cc6b3c468
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f581254c065ba540cc0ee25a8155b028c066cf5c2d85c7ea62849c1cf9096b4e
+size 338411
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..a8563ab48869fa9901ea2a5ee0cc6ea6f7c4d912
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a chef named Timothy Grandinetti, who is seen in a professional kitchen setting. He is wearing a black chef's coat and glasses, and he is holding a bowl in his hands. The kitchen is equipped with a large oven and a grill, suggesting that it is a commercial or professional kitchen. The chef appears to be in the middle of a cooking demonstration or presentation, as he is gesturing with the bowl and seems to be explaining something. The overall style of the video is informative and educational, likely aimed at teaching viewers about cooking techniques or recipes."
+    ],
+    "video_ids": [
+        "CUSHf-cPFc8_0_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chef named Timothy Grandinetti, a black chef's coat, glasses, a bowl? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows Chef Timothy Grandinetti wearing a black chef's coat and glasses, standing in front of a bowl. All core elements from the description are present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a Timothy Grandinetti wearing a black chef's coat and glasses. which in a of a kitchen. The the elements of the description are present, accurately depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e967839022649d57416e59b76c2b7f67a25f72fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d4cbd8f536b3e9bf25b13c3b052d1f1a72cd5bcd6d832763663012484a738be0
+size 157951
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..b7a64da73a5ff35a29c4b2f1f2043b1351d5baaf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment on a soccer field. A soccer player, dressed in a vibrant blue and red striped jersey, is in the midst of a triumphant celebration. His arms are outstretched, embodying a sense of victory and joy. The player is the central figure in the image, with the soccer field stretching out behind him. The field is a blur of green, indicating the high-speed action of the game. The player's jersey is clearly visible, with the word \"Rakuten\" prominently displayed. The overall style of the video is energetic and exciting, capturing the essence of the sport and the passion of the players."
+    ],
+    "video_ids": [
+        "G35kAJVk8F4_33_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A soccer player in a vibrant blue and red striped jersey displaying the word 'Rakuten'.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a vibrant blue and red striped jersey with the word 'Rakuten' clearly visible on the front, matching the description. The player's pose and the background context (stadium, crowd) are consistent with a soccer match setting, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a vibrant blue and red striped jersey with the word 'Rakuten' prominently visible on the front. which the description provided The player is attire and the setting of ofaadium setting grass) also consistent with a soccer match setting, further the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5314e008ceb9bc69d5ac98d60e9380645769ce41
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1333-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ffcef04631a4801da4f64c33736f5f1c3e76b603b93c7649a329e350910fa76
+size 137797
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..3b0d5439dcceca23ee849fcd021d855fffd5d61d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red sports car in motion, captured in three frames. The car is sleek and low to the ground, with a shiny finish that reflects the environment. The first frame shows the car from a side angle, with the driver's side door open, revealing the interior. The second frame shows the car from a rear angle, with the license plate visible. The third frame shows the car from a front angle, with the hood up, revealing the engine. The car is on a gray surface, possibly a track or a road. The style of the video is dynamic and high-speed, capturing the car's movement and design in a way that emphasizes its speed and power."
+    ],
+    "video_ids": [
+        "8AnOBe11ppM_30_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red sports car, specifically a Ferrari F50, which matches the description. The car is shown from various angles, highlighting its red color and sports car design. Additional elements in the background (like other cars or garage elements) do not contradict the core description of a red sports car.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a red sports car in which a Tesla,80, which is the description of The car is shown from multiple angles, highlighting its sleek color and sleek car design. The elements such the video,like the cars and tracks settings) do not contradict the core description of the red sports car.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5d05ed37b0d073fe37dff51c4769564de7d865cc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0d81a77660c640b1fb32a99cf98d4a3f22480662affbaf1c9e56908a27ac01ea
+size 278635
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..28ff55d84d4c89eaa07eb940ec11b3261c282630
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a blue jacket opening the hood of a red car in a grassy field. The man is standing to the left of the car, looking down at the engine. The car is parked on a grassy area with trees in the background. The man appears to be examining the engine, possibly checking for any issues or maintenance needs. The scene is set in a natural outdoor environment, suggesting that the car may have broken down or the man is performing routine maintenance checks. The overall style of the video is realistic and documentary-like, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "xUOhJux8iEc_11_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue jacket, a red car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a blue jacket standing next to a red car with its hood open. The man is interacting with the car, and the red car is prominently featured throughout the video. There are no elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man in a blue jacket and next to a red car with its hood open. The man appears interacting with the car, which the setting car is the featured in the frames. The are no elements in contradict the description description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ed930a91be4aba75f9c7c5c2571e978779177bbd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb5b99911d26f5f49977e79f224712ae59e227516533cc32aa621656bdf2abab
+size 173849
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..653f326a835d05e8903fa2dab1f9ec91474b6212
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a promotional advertisement for hammered copper barware. It features a close-up of a copper pitcher and a copper cup, both with a textured, hammered finish. The copperware is placed on a wooden cutting board, which also holds fresh limes and mint leaves, suggesting a refreshing beverage. The background is blurred, but it appears to be a kitchen or bar setting. The style of the video is clean and modern, with a focus on the copperware's texture and the fresh ingredients, likely to evoke a sense of luxury and sophistication. The video is likely to be used in a catalog or online store to showcase the product's design and potential use."
+    ],
+    "video_ids": [
+        "5UXIzzeGKaY_1_0to185"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Copper pitcher, copper cup, wooden cutting board, fresh limes, mint leaves? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully features a copper pitcher (shaker), copper cups (mugs), a wooden cutting board with mint leaves, and fresh limes. These core objects are clearly visible and match the description. Additional elements like a jigger and a tray are present but do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a copper pitcher,bothiny), a cup,mugs), a wooden cutting board, fresh leaves, and fresh limes. The elements elements are present visible and match the description provided The elements like the secondigger and a blurred are not but do not contradict the core and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6b9dc2718b0590c06f11dd1f8bdc41fb444b87ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7cf9785e8ad18203064237dada1ae2743a6cc230e521094c189c3bf409b3e20e
+size 115316
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..026236d958357b3a7080af8892bf4180b53dd07b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a car and a motorcycle parked in a garage. The car is a small, compact vehicle with a distinctive front grille design. The motorcycle is parked next to the car, and it has a sleek, modern design. The garage has a concrete floor and white walls, and there are colorful paintings on the walls. The lighting in the garage is bright, and there is a light on the ceiling. The car and motorcycle are stationary, and there is no movement in the video. The style of the video is straightforward and documentary, with no special effects or artistic filters. The focus is on the car and motorcycle, and the garage serves as a simple, uncluttered background."
+    ],
+    "video_ids": [
+        "3g0hc0pAwD8_53_23to144"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small, compact car with a distinctive front grille and a sleek, modern motorcycle.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a small, compact car with a distinctive front grille, which matches the description. Additionally, a sleek, modern motorcycle is visible in the background, also aligning with the description. The presence of other elements like bicycles and clothing does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a small, compact car with a distinctive front grille, which matches the description. Additionally, there sleek, modern motorcycle is visible in the background, also aligning with the description. The setting of the elements like the and the in not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1aba8f5947e74fcb6b1b56483d5a36c212991605
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d778ae437c073cff45a0024a8b8ac3b4ff2b3bfa63819956a27b54f656b2bf8a
+size 229133
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..56881f91fc0111d62be4281e0893b0dd91f2b600
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are standing in front of a brick wall. The woman has pink hair and a tattoo on her arm, while the man is wearing a colorful hat. They are both looking at something off-camera, and the woman is holding her hair back. The video has a casual and relaxed vibe, with the two individuals appearing to be in a good mood. The brick wall in the background adds a rustic touch to the scene."
+    ],
+    "video_ids": [
+        "UvdBm9R0Jbg_8_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman standing in front of the brick wall; the woman has pink hair and a tattoo, the man wears a colorful hat; the woman is holding her hair back; both are looking off-camera.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a woman standing in front of a brick wall. The woman has pink hair and a visible tattoo on her arm, and she is holding her hair back with a clip. The man is wearing a colorful hat. Both individuals are looking off-camera, which matches the description. The video does not contradict any of these core elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a woman standing in front of a brick wall. The woman has pink hair and a tattoo tattoo on her arm, and the is holding her hair back with her hand. The man is wearing a colorful hat and Both individuals are looking off-camera, which align the description. The presence does not contradict the of the elements elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0a7df169702930ba5e789e3812253862f043a7c3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:29f1d09156d238f4a7082a27dd0c2f667f52f1bf3a51c15ad522eb8db0ef73ba
+size 223939
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..79f0a3b399c040805eb24d63210833b204b9299c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a friendly brown bear is seen in a lush garden filled with vibrant red berries. The bear, with its brown fur and a cheerful expression, is seen interacting with a basket of berries. In the first frame, the bear is seen approaching the basket, its eyes focused on the ripe berries. In the second frame, the bear has reached the basket and is carefully picking up a berry with its paws. In the third frame, the bear is seen holding the berry in its mouth, ready to enjoy the sweet treat. The garden around the bear is lush and green, with the red berries adding a pop of color to the scene. The bear's actions and the surrounding environment create a peaceful and serene atmosphere."
+    ],
+    "video_ids": [
+        "M0r-s8tr1Cw_51_0to140"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A friendly brown bear, a basket of berries? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a friendly-looking brown bear holding a basket filled with berries, which matches the described 'Object(s)'. The bear is actively engaged in picking berries, and the basket is clearly visible. Additional elements like the garden setting do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a brown brown brown bear and a basket filled with red. which align the description 'Object(s)'. The bear is interacting interacting with the berries, and the basket is clearly visible, The elements like the lush setting and not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..49a26a88ebe5d2751bba9ed1d525a5b70cfc6d35
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b79d41e0a5f23a73d771022a0b2d88e0e9c9d8bf7c7cd8d51ed6ed6d60478b7
+size 262774
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..256226172ddaecd7c02fff1a3921e9d4a93b115f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a white car driving on a sandy beach at sunset. The car, with its brake lights on, is the main focus of the image, moving from the left to the right side of the frame. The sky, painted in hues of orange and pink, serves as a stunning backdrop to the scene. The sun is setting in the distance, casting a warm glow on the landscape. The vast expanse of the beach stretches out in the background, with a few hills visible in the distance. The overall style of the video is calm and peaceful, capturing the beauty of a quiet moment on the beach at sunset."
+    ],
+    "video_ids": [
+        "2hUYzR5M-FU_49_0to188"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a white car, specifically a Cadillac CTS station wagon, which is the central object in the scene. The car is clearly visible throughout the video, and its color and model are consistent with the description. The presence of a sunset and desert landscape does not contradict the core description of a white car.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a white car, which a white,TS-V wagon, which is the main object in the scene. The car is positioned visible and the frames, parked its white and model are consistent with the description. The setting of the sunset and a landscape in not contradict the ' description of the white car.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d72265e6b627b6f8e7095fdaa932597bd6e43ee3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3cabc3fc210e3c5d3679d043a2974d5a6b6f240dcb57348b553a9045f04eb90e
+size 211085
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..7669966ff8df3fafb8ae3b770850c580902dff96
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a humorous and informal vlog. A man with a beard and a surprised expression is the main subject. He is shirtless and pointing at a picture of another man holding a book. The picture is superimposed on the video. The background shows a cozy living room with a bookshelf, a red couch, and a lamp. The man in the picture is holding a book titled \"Sing 4\" and appears to be excited or surprised. The overall style of the video is casual and personal, with a focus on the interaction between the two men and the book."
+    ],
+    "video_ids": [
+        "U-A2GhOjL5Q_10_219to388"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and a surprised expression (main subject), a picture of another man holding a book.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The main subject is a man with a beard who appears to have a surprised expression, matching the description. Additionally, there is a picture of another man holding a book (specifically, the 'Sing 4' game box) in the top right corner of the frame, which also matches the description. No elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video subject in a man with a beard and appears surprised be a surprised expression. which the '. Additionally, there is a picture of another man holding a book,whichally, a bookF'1' book box), which the video right corner, the frame, which align align the description. The other contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2d2473636ac0f4093c7b0c1de475059583c3cf5c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1334-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:074c563c4a884d1859121fb760263db9fba414daca459d858fbf9790e83e8e9b
+size 219166
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..6f5ed19d6b62e3318613671e8363ec5c7014a5c5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two young men are engaged in a conversation in a room with a window. The man on the left is wearing a gray beanie and a plaid shirt, while the man on the right is dressed in a purple sweater. They are seated on a couch, with the man on the right holding a guitar. The room has a cozy atmosphere, with the window allowing natural light to enter. The men appear to be engaged in a serious discussion, as they look at each other intently. The guitar suggests that they might be musicians or music enthusiasts. The overall style of the video is casual and intimate, capturing a moment of connection between the two men."
+    ],
+    "video_ids": [
+        "RUcLuQ17UV8_146_94to253"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two young men, a couch, a guitar? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two young men sitting on a couch, with one of them holding a guitar. The scene matches the described objects without contradiction, even though additional elements like a window and wall are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two young men sitting on a couch. with one of them holding a guitar. The presence matches the description elements without any.\"\n fulfilling though the elements like the window and a are present,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..70b1d4b6664f85493406f521396fccf94a067176
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:61f1157f5d8e9bfd99a9e85b3dc31cc7f7d27bb55e75d06481bfc733fdb53b49
+size 183281
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..53d38a6182a6c0af01d2fde5d689780762695c5a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person's hand is seen holding a small, wooden display case containing miniature cakes. The display case is placed on a white surface, and the cakes inside are of various colors and designs. The hand is in the process of opening the display case, revealing the cakes inside. The background features a miniature model of a bakery, complete with a counter and shelves filled with more miniature cakes. The overall style of the video is a close-up, slow-motion shot that captures the details of the miniature cakes and the display case. The focus is on the display case and the cakes, with the bakery model serving as a backdrop. The video does not contain any text or additional elements."
+    ],
+    "video_ids": [
+        "Ss1dUoeULqc_68_0to116"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person's hand, a small wooden display case, miniature cakes? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person's hand interacting with a small wooden display case filled with miniature cakes. The hand is visible placing and adjusting the display case, and the miniature cakes are prominently displayed inside the case. The background elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a person's hand interacting with a small wooden display case. with miniature cakes. The display opens seen as the opening the lid case, and the miniature cakes are distinctly displayed inside the case. The video also, not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d548f49bfdbee3d15012cc8b629498f4393515ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:deeb61f40e5ad318085b722ebcc311229914c4fc793c9feb66c9b9d82b6d1769
+size 115292
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..7c8a3ce8c273c1c252fea205dfe3b5e43e6bf96f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a chef named Timothy Grandinetti, who is seen in a professional kitchen setting. He is wearing a black chef's coat and glasses, and he is holding a bowl in his hands. The kitchen is equipped with a large oven and a grill, suggesting that it is a commercial or professional kitchen. The chef appears to be in the middle of a cooking demonstration or presentation, as he is gesturing with the bowl and seems to be explaining something. The overall style of the video is informative and educational, likely aimed at teaching viewers about cooking techniques or recipes."
+    ],
+    "video_ids": [
+        "CUSHf-cPFc8_0_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chef named Timothy Grandinetti, a black chef's coat, glasses, a bowl? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows Chef Timothy Grandinetti wearing a black chef's coat and glasses, standing in front of a bowl. All core elements from the description are present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a Timothy Grandinetti wearing a black chef's coat, glasses. which in a of a kitchen. The the elements of the description are present, accurately depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..432eeeeb621967860ad8bb508a94f093e0ed0c42
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4878d31f045549a84382e818aa2a1987a2bf0d2715aa7931804c75bee383ab53
+size 157003
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..2039bfbf52ec67f2cc15225da5c91136a8afc6ce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman and a young girl are seen in a kitchen, engaging in a daily activity. The woman, dressed in a blue shirt, is standing at the sink, washing dishes. The young girl, wearing a green dress, is standing on a stool next to the woman, observing and learning. The kitchen is well-equipped with wooden cabinets, a sink, and various kitchen items such as a kettle, a bowl, and a bottle. The scene captures a moment of bonding and learning between the woman and the child, set in a warm and inviting kitchen environment."
+    ],
+    "video_ids": [
+        "I62TeuSuMPg_1_0to147"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a blue shirt and a young girl in a green dress.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a blue shirt and a young girl wearing a green outfit (appears to be a green dress or long-sleeved top with polka dots) standing at a kitchen sink. The core description is accurately represented, and no elements contradict this.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a blue shirt and a young girl in a green dress.aars to be a dress dress). a-sleeved top). greenka dots). standing in a kitchen sink. The woman elements of accurately represented in with there additional contradict the.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3ff0add3ae19220a7d0604d5fc3283f6f7aff3d2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:102b9845f3cf69456086ce84253f64dfe7f7f0538dbddeb64b23f1b2840a57c2
+size 159354
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..10f80c3ee5a4fa028937bb337e42534e41a5b2d9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the majestic Angel Falls, the world's highest uninterrupted waterfall, cascading down a towering cliff in the heart of the Amazon rainforest. The waterfall's powerful flow creates a misty veil that shrouds the cliff face, adding a mystical quality to the scene. The lush greenery of the surrounding rainforest provides a vibrant contrast to the waterfall's white cascade. The video is taken from a distance, allowing the viewer to appreciate the waterfall's grandeur and the vastness of the rainforest. The perspective of the video emphasizes the waterfall's height and the scale of the natural landscape. The image is a testament to the beauty and power of nature."
+    ],
+    "video_ids": [
+        "U7sGxKoCjRo_76_305to474"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Angel Falls (waterfall), towering cliff, misty veil, lush greenery? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts Angel Falls, the tallest uninterrupted waterfall in the world, cascading down a towering cliff. A misty veil is visible at the base of the waterfall and enveloping the surrounding landscape. The foreground and midground are filled with lush greenery, including dense forests and trees, matching the described elements. There are no contradictions with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful Angel Falls, a towering waterfall waterfall in the world, whichading down a towering cliff. The misty veil is visible at the base of the falls, throughouting the area area, The lush is backgroundground are filled with lush greenery, which dense forests and vegetation, which the description '. The are no contradictions in the given description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c889badf9a40ad509cc7fa51e7998858a9f9e41a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d298a4513033e376433421a12b8738c85d7394d6813e12f201948fd5690d75c2
+size 178739
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..c09f3b553791943bbe52500e09cd5be5b0b00513
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases the interior of a luxury car, specifically a Mercedes-Benz, in three different angles. The first angle is a close-up of the steering wheel, which prominently displays the Mercedes-Benz logo. The second angle provides a view of the dashboard, featuring a sleek design with a touch screen display. The third angle offers a perspective of the passenger side, highlighting the plush red leather seats and the door panel with controls and a window. The overall style of the video is sleek and modern, emphasizing the car's luxurious features and design."
+    ],
+    "video_ids": [
+        "-7bAe7gvGrw_4_0to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, dashboard, passenger side seats, door panel with controls and window.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the steering wheel with the Mercedes-Benz logo, the dashboard with gauges and infotainment screen, the passenger side seats upholstered in red leather, and the door panel with controls and window. All elements described in the condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows the steering wheel, the Mercedes-Benz logo, the dashboard with auges and aotainment system, the passenger side seats,stered in red,, and the door panel with controls and window. The these match in the ' are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..875aa6d52111953725b1d687d7b5368f9383c612
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fed7e590248cd4a6dc6c14a87b160dce9be7418b491134c80b26b5f06bd3b203
+size 136682
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..85bf1c2283ea1ec656552ac973add89c7d4abe54
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a rugged off-road vehicle, likely a Jeep, driving through a muddy field. The vehicle is covered in mud, indicating it has been off-roading or traversing rough terrain. The field is surrounded by tall, dry grass and trees, suggesting a rural or wilderness setting. The vehicle is moving forward, and the mud splashes up around its tires as it moves. The style of the video is realistic and captures the vehicle's movement and the surrounding environment in detail. The focus is on the vehicle and its interaction with the muddy field, with no other significant objects or characters in the scene. The lighting is natural, and the colors are muted, with the mud and the green of the grass being the most prominent. The video does not contain any text or additional graphics."
+    ],
+    "video_ids": [
+        "Gnf2d0LyHXM_33_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A rugged off-road vehicle, likely a Jeep.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a rugged off-road vehicle, clearly identifiable as a Jeep Wrangler, covered in mud, navigating through a muddy, grassy terrain. The vehicle's design, including its boxy shape, spare tire mounted on the rear, and overall appearance, matches the description of a rugged off-road Jeep.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a vehicle off-road vehicle that which resembling as a Jeep,angler, navigating in mud, which through a natural terrain grassy terrain. The vehicle's design, including its boxy shape, large tire mounted on the back, and the structure, align the typical of a Jeep off-road vehicle.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..57c664cf71e7b6adf8c751933f9c64060774f029
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9cd1670dbee8dddcb771eaf6a4765f3dc711f4887b089a36c2aafe2cfcff9196
+size 407667
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..8f871311a051317b1963c7dfa7d89b22a9043314
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a vintage car in a field, with a focus on the rear end. The car has a distinctive chrome bumper and tail fins, which are rusted and show signs of wear. The car is parked on a dirt field, with trees and bushes in the background. The video is shot in a realistic style, capturing the details of the car and its surroundings. The car appears to be in a state of disrepair, suggesting it has been abandoned or neglected for some time. The overall mood of the video is nostalgic and melancholic, evoking a sense of the passage of time and the decline of the car's once-glorious appearance."
+    ],
+    "video_ids": [
+        "ReZPLsInAhA_16_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Vintage car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a vintage car, specifically an old, rusted model with classic design elements such as tailfins and chrome details. The car's aged and weathered condition aligns with the 'vintage' description, and no elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features the vintage car, specifically focusing older model classiced model with a design elements such as a finsins and a details. The setting is condition appearance weathered appearance,s with the 'vintage' description. and the other contradict this core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8d9c5d0756ee77f20182b217de33b2d198cf8467
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1335-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:598d1f124003ebc91370c3dc9fbb1adec3a180835888a7f234c267d292a5f690
+size 173977
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..13db1f3ca84ab462eb382119c4cc44debf960c16
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a woman sitting in a restaurant booth, smiling at the camera. She has long blonde hair and is wearing a white t-shirt. Her nails are painted blue and she is wearing multiple rings on her fingers. The restaurant has a yellow and green color scheme and there are other patrons in the background. The style of the video is casual and candid, capturing a moment in the woman's day."
+    ],
+    "video_ids": [
+        "XF7fvvUEl2A_64_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman sitting in a restaurant booth, smiling at the camera. She has long blonde hair, is wearing a white t-shirt, and has blue-painted nails with multiple rings on her fingers.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman sitting in a restaurant booth, smiling at the camera. She has long blonde hair, is wearing a white t-shirt, and has blue-painted nails with multiple rings on her fingers. These details align with the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman sitting in a restaurant booth, smiling at the camera. She has long blonde hair, is wearing a white t-shirt, and has blue-painted nails with multiple rings on her fingers. The elements match with the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..22333c1cd020688a0a86ac1aaaf6a46650cacd75
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41f29f677edeb0055f1e3fa87597d4850f04b7fa9defee444178d4f7e98b2a6c
+size 97644
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..1fd62904c19435c57104a33d19f029b4d55f4c2c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a bustling city street scene. The sun is shining brightly, casting a warm glow on the scene. The street is lined with tall buildings, including a prominent one with a clock tower. The traffic lights are visible, indicating a busy intersection. Cars are driving down the street, adding to the urban atmosphere. The street is also adorned with trees, providing a touch of nature amidst the concrete jungle. The overall style of the video is a blend of urban and natural elements, creating a dynamic and vibrant cityscape."
+    ],
+    "video_ids": [
+        "kYX1DDxjkoc_52_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Tall buildings, a building with a clock tower, traffic lights, cars, trees? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows tall buildings, including one with a clock tower (the central building), traffic lights (visible on poles), cars (multiple vehicles on the road), and trees (green foliage around the area). All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful tall buildings, a one with a clock tower,the clock tower in traffic lights,visible at the), cars (multiple vehicles on the road), and trees (liningery lining the buildings). The these elements of in present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..23fbd907e6905fd1b8d1b994c015238c2c00dc65
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ad664eb4f439aa8bf8c764ea2216bd6f9a028842f3799c7fdc2d6d823585a31f
+size 250681
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..6b6e430081af4c6debd171f00b08b5b298385c70
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a luxurious wooden cabin with a cozy and inviting atmosphere. The cabin features a warm wooden interior with a large window that offers a stunning view of a serene lake. The cabin is furnished with comfortable seating, including a plush couch and a cozy chair, both adorned with soft cushions. A wooden dining table is positioned in the center of the room, surrounded by chairs, inviting guests to enjoy a meal or conversation. The cabin's design and furnishings create a sense of relaxation and comfort, making it an ideal retreat for those seeking a peaceful escape. The video captures the essence of the cabin's charm and the beauty of the surrounding nature, offering a glimpse into a tranquil and idyllic getaway."
+    ],
+    "video_ids": [
+        "Djb5dyKRISs_59_0to174"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Plush couch, cozy chair, wooden dining table, chairs? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a wooden dining table with a central pedestal base, surrounded by cushioned seating that resembles plush couches and cozy chairs. The seating is upholstered in light fabric with pillows, fitting the description. The wooden table and surrounding seating are clearly visible and central to the scene, confirming the core objects are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a cozy dining table with chairs plush table,, surrounded by chairsed chairs that includes a couches and cozy chairs. The room arrangement arrangedstered in a-colored, brown, which the description of The wooden dining and chairs furniture are consistent visible, match to the scene, fulfilling the presence elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0013e6bc555ba6e0a564ebf1cd2047a825a65330
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60a1017bb5095384c9f76d34337d5fbc76e46d8131915695abd2abd5ba95fd13
+size 234328
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..668c2539c84c2570c79aa5c2f3961d3e02f9b9d5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a black shirt standing in front of a yellow building with a blue sign that reads \"JERK CENTRE\". He is wearing sunglasses on top of his head and has a slight smile on his face. The man is looking to his left, and his neck is slightly tilted. In the background, there are two women standing near the entrance of the building. The style of the video is casual and candid, capturing a moment in the man's day."
+    ],
+    "video_ids": [
+        "a6F4vuYPpi8_84_86to293"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black shirt, two women near the building\u2019s entrance.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a black shirt as the central subject, and two women are visible near the building\u2019s entrance in the background. These elements align with the described condition, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black shirt and described main figure, and there women are visible near the building's entrance in the background. The elements match with the description ', making there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9aa7dbaf6266b80f567010739c73da59c8bacce3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3072fd77c30bdbb7c4bfd9e9abc660b8bc70aa4c052049a957389ca2a3f31f17
+size 146284
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..c1f1548ab2df15f171004bda89a2bf33f5163a27
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the breathtaking view of a waterfall cascading down a rocky cliff. The waterfall, with its white, frothy water, is the central focus of the video. The rocky cliff, with its rugged texture, provides a stark contrast to the smooth flow of the waterfall. The waterfall's position on the cliff creates a dynamic interplay of light and shadow, adding depth to the scene. The video is taken from a high angle, providing a bird's eye view of the waterfall and the cliff. This perspective allows the viewer to appreciate the scale and grandeur of the waterfall and the cliff. The video is a testament to the beauty of nature and the power of water."
+    ],
+    "video_ids": [
+        "wvvQ8PeHH5g_81_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Waterfall, Rocky Cliff? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a waterfall cascading down a rocky cliff, matching the specified objects. The surrounding terrain is rugged and rocky, consistent with a cliff environment. No conflicting elements are present that contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a waterfall cascading down a rocky cliff. which the ' '. The presence environment and consistent and consistent, and with a cliff,, The additional elements are present.\"\n would the core.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0febafc44cc18ec5e56139442437effc586fe21d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d05b3d974550dcdae7ac86f1643eff350616610d16fff9792707cfc48c3ee06
+size 230566
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..cf6bb9476d7b7d3577dfa0bd1380ea1f2e9ec935
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up, slow-motion shot of a golden chalice being placed on a golden tray. The chalice is intricately designed with a handle and a lid, and it is placed on a tray that has a lid and a handle as well. The tray is placed on a table, and there is a golden vase next to it. The background is blurred, but it appears to be a room with a window. The lighting in the room is soft and warm, highlighting the golden objects. The video captures the elegance and sophistication of the objects, and the careful placement of the chalice on the tray."
+    ],
+    "video_ids": [
+        "lcT4nz-TLVo_28_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A golden chalice and a golden tray with lids and handles, a golden vase? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features golden chalices, a golden tray with intricate designs (which can be interpreted as having lids and handles), and a golden vase with handles. These objects are clearly visible and match the description. The presence of other golden items in the background does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features golden objectsalices, trays golden tray with lids designs,which could be interpreted as having lids and handles), and a golden vase. a. The objects match consistent visible and match the description provided The video of a golden objects in the background does not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a03e4d3d5cc88c32bc419f9239cd55a8255643eb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e3bc618fb0bd0a86bb103d691cd1e00120a614dea215376ba5605eea7c83159
+size 250934
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..55efee21a77f7a3f02e4e08246ab97c6a4eea735
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young man and woman are seated at a table, engaged in a playful activity. The man is holding a small potato in his hand, while the woman is writing in a notebook. The table is adorned with a vase filled with colorful pencils, adding a vibrant touch to the scene. The backdrop features a bookshelf filled with books, suggesting a studious or intellectual setting. The overall style of the video is casual and light-hearted, capturing a moment of fun and creativity between the two individuals."
+    ],
+    "video_ids": [
+        "RJwB-agReSU_17_88to231"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man holding a small potato, a woman writing in a notebook? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man holding a small potato (which appears to be a cookie or similar object) and a woman writing in a notebook. These core elements are present and consistent with the description, even though there are additional decorative and playful elements in the scene (e.g., crayons, green antennae, colorful background).\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man holding a small potato andwhich could to be a golden or a object) and a woman writing in a notebook. The elements elements match present and match with the description. even though the are additional elements elements colorful elements like the video,such.g., colorfulons and bookseryae, and background). The}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d5082eaf13d6d241d8a239f1de46f4cf2fe87c7c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a4629e5f3a35c1902c2b91ea9f38bbafa39b0a4727459f669546d93727ef755
+size 165529
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..ea3ff389427839d5d9147af3823fc2bee0b15fc7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and short hair, wearing a striped shirt. He is seated on a couch in a room with a brick wall and a colorful mural in the background. The man appears to be engaged in a conversation or interview, as suggested by his attentive expression and the presence of a microphone clipped to his shirt. The overall style of the video is casual and relaxed, with a focus on the man and his surroundings. The lighting in the room is soft and warm, creating a comfortable atmosphere. The brick wall and mural add a touch of urban art to the setting."
+    ],
+    "video_ids": [
+        "5nNrTkrDfT4_3_24to223"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and short hair, wearing a striped shirt, seated on a couch. A microphone clipped to his shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and short hair, wearing a striped shirt, seated on a couch. A microphone is visibly clipped to his shirt. All elements of the description are accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and short hair, wearing a striped shirt, seated on a couch. There microphone is clipped clipped to his shirt, The elements of the description are present represented in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..474570356d50268c5a5776399c33c61d8d1b3152
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1336-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fc03459a09f60369ffc4a50568f7256ae764910b55c1a9884a85eaa7f6e98d66
+size 199767
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..672ea16bbb83bd6d02409edc2c45f43b146fa645
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a white SUV parked inside a car dealership. The car is positioned in the center of the frame, with its front facing the camera. The dealership has a large window that offers a view of the outside, where other cars are parked. The interior of the dealership is well-lit, with a clean and polished floor. The car is the main focus of the video, and it appears to be in excellent condition. The dealership itself is spacious and well-organized, with other cars on display in the background. The video is likely intended to showcase the SUV for potential buyers, highlighting its design and features."
+    ],
+    "video_ids": [
+        "_zd0HzDyyGc_8_0to172"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white SUV? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a white SUV, specifically a BMW X3, which is the main subject of the footage. The camera pans around the vehicle, clearly showing its white exterior and other details, fulfilling the 'Object(s)' condition of showcasing a white SUV.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a white SUV, which a Lincoln X5, which matches the main object. the frames. The vehicle angle around the vehicle, showcasing showing its design color and design features such such the 'Object(s)' condition of the a white SUV.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f04cb740cc8b77ffe2e4f51b79a06528b19c444f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:662d44c964e414a139a5308ff68a97ac9624b774a374cc312b3063e2cd57dd05
+size 159599
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..123b35b6874b5ae8eb02109c75ed926c98f6d4e2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man wearing glasses, looking down with a serious expression. The man is dressed in a light-colored shirt and appears to be in an indoor setting with a blurred background. The style of the video is a close-up shot, focusing on the man's face and upper body. The lighting is soft, and the overall mood of the video is contemplative."
+    ],
+    "video_ids": [
+        "JUAx3IGN0aE_82_0to182"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man wearing glasses, looking down with a serious expression, dressed in a light-colored shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a man wearing glasses, looking downward with a serious expression, and dressed in a light-colored shirt. The description matches the visual content without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man-up of a man wearing glasses, looking down with a serious expression. and dressed in a light-colored shirt. The description matches the visual elements of any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8956aff0a98454a106cd74f0778d39e5a1a7fbaa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f1a665c6092c2646f514a3e18bcb862343a119860af843b6fe9b71908488f25
+size 77038
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..55e6d70870c8fdd843442158b343b28eb5e9350e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a bustling city street scene. The sun is shining brightly, casting a warm glow on the scene. The street is lined with tall buildings, including a prominent one with a clock tower. The traffic lights are visible, indicating a busy intersection. Cars are driving down the street, adding to the urban atmosphere. The street is also adorned with trees, providing a touch of nature amidst the concrete jungle. The overall style of the video is a blend of urban and natural elements, creating a dynamic and vibrant cityscape."
+    ],
+    "video_ids": [
+        "kYX1DDxjkoc_52_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Tall buildings, a building with a clock tower, traffic lights, cars, trees? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows tall buildings, including one with a clock tower (the central building), traffic lights (visible on poles), cars (multiple vehicles on the road), and trees (green foliage around the area). All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful tall buildings, a one with a clock tower,the church tower in traffic lights,visible at the), cars (blue cars are the road), and trees (liningery on the buildings). The these elements of in present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..24f7e9e56743c816b7f4c6080b0575c47b27de59
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4076f4d6fba41c89cd5b8de5511dd8522445220eaf9729210b1f35fb5f04cb1c
+size 222006
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..fa59d8dfa687097f6d9a6431ed743eafa6f3a03d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a scene on a city street where a silver sports car is parked. The car is sleek and modern, with a low profile and a shiny finish. It's parked on the side of the street, and the camera angle is from the front, giving a clear view of the car's design.  As the video progresses, two men approach the car. They are casually dressed, with one wearing a blue shirt and the other in a white shirt. They seem to be admiring the car, possibly discussing its features or simply enjoying the sight.  The background of the video shows a typical city street with parked cars and buildings. The street is lined with trees, adding a touch of nature to the urban setting. The overall style of the video is realistic, capturing a moment in time on a city street."
+    ],
+    "video_ids": [
+        "MAmMk82ni7c_47_0to162"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A silver sports car, two casually dressed men (one in a blue shirt, one in a white shirt).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a silver sports car, specifically a Chevrolet Corvette, parked on a street. There are two casually dressed men visible: one wearing a blue shirt and another in a white shirt. Their presence and attire match the description, and no conflicting elements are present that contradict the core requirement.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a silver sports car parked which a sleek Corvette, which on the street. A is two men dressed men, in one in a blue shirt and the wearing a white shirt. The presence and casual match the description. and the additional elements are present.\"\n would the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4ce2ddf880037871532e974c80b9f4db794cb62d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30da064da02bc85e19b265f50ce4aba992ec76bf00c8e432a8a7645eaf465cfa
+size 224060
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..cd3aa12dd92ebc8594082b072261289e925dc007
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of triumph and camaraderie among a group of soccer players. The players, all dressed in vibrant yellow jerseys, are seen celebrating a goal on the field. The first frame shows a player in the foreground, his arms raised in a victorious gesture, while his teammates in the background are seen running towards him, their faces lit up with joy. The second frame captures the same moment from a different angle, with the celebrating player now in the background and his teammates in the foreground, their bodies in motion as they join the celebration. The third frame is a close-up of the same player, his face beaming with pride and his teammates' hands on his shoulders, a symbol of their shared success. The video is a dynamic and colorful depiction of teamwork and sportsmanship."
+    ],
+    "video_ids": [
+        "TYmChAG9pfI_27_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Soccer players, a player in the foreground, teammates in the background, and another close-up of the same player with teammates.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple Brazilian soccer players in yellow jerseys celebrating on the field. The camera pans across them, highlighting a player in the foreground (number 16) and then revealing teammates in the background. The sequence includes a close-up of the same player (number 16) with teammates around him, fulfilling the described conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a frames soccer players in yellow jerseys, on a field. The first captures to the, focusing a player in the foreground withlikely 10) with teammates focusing teammates in the background. The sequence also a close-up of the same player withnumber 16) with teammates, him, which the ' conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..88836f0eec680997e4a55209a9a096360d63a823
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:080cc22af0bef8a5759b123a94eb9dd4ef1c178c5b3ab0dfea6fa7654a4fd07f
+size 246942
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..f1600ae2bb0c524751277b98d87c975b03243f49
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a modern living room with a large glass door leading to a balcony. The room is well-lit with natural light from the balcony. The living room features a comfortable couch with a coffee table in front of it. A flat-screen TV is mounted on the wall, and a rug is placed on the floor. The room is furnished with a dining table and chairs. The balcony offers a view of the city skyline. The video captures the essence of a contemporary living space with a blend of comfort and style."
+    ],
+    "video_ids": [
+        "DiDuMc8E7BI_22_56to194"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Comfortable couch, coffee table, flat-screen TV, rug, dining table, chairs? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a comfortable couch with cushions, a coffee table in front of it, a flat-screen TV mounted on the wall, a rug under the coffee table, and a dining table with chairs visible in the background. All specified objects are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a living couch, a, a coffee table, front of it, a flat-screen TV mounted on the wall, a rug on the coffee table, a a dining table with chairs visible through the background. The the objects are present and match with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2461555724d48c1d6a0c195866b91eed31b9bd77
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c8b9ab32b0f6984c6bb1ea9f435ce5c05f5036554bcdc424a837392a5142d7ec
+size 126644
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..3e568f4f934d9374bc0b0416a4f134a0ff72cc7c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with a colorful makeup look, including vibrant eyeshadow and lipstick. She is wearing a headscarf and a gray jacket. The background consists of a shelf with various items, including books and boxes. The woman appears to be speaking or making a statement, as she is looking directly at the camera with her mouth slightly open. The overall style of the video is casual and personal, possibly a makeup tutorial or a personal vlog."
+    ],
+    "video_ids": [
+        "EEgSzMET_FI_9_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with a colorful makeup look, including vibrant eyeshadow and lipstick, wearing a headscarf and a gray jacket. The woman is speaking or making a statement.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with vibrant, colorful eyeshadow (pink and blue/green hues) and lipstick, wearing a headscarf and a gray jacket with a leopard print pattern. She is looking at the camera and appears to be speaking or making a statement, as indicated by her mouth movements and direct gaze. The background, though blurred, does not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with a makeup colorful makeupeshadow andthough and orange tonesreen tones) and lipstick ( wearing a headscarf with a gray jacket. a black print design. The appears speaking directly the camera, appears to be speaking or making a statement, which her by her open movements and the gaze. The background includes while not, suggests not contradict the description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..28d5f8f3a73958ee2d023d92f168c0c62b2ab161
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f818a4385056361b5cc5e0dcf5076968ac66934226cd90763e843403e3ca4cbc
+size 239840
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..fd8cc8e6d1943a1a36972a987a796fb5d1b78af7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a chef in a white uniform standing in front of a large stainless steel grill. He is holding a tray with three pieces of food on it. The chef is smiling and appears to be presenting the food. The setting is outdoors, with a wooden deck and a fence visible in the background. There are also some chairs and tables on the deck. The lighting suggests it is daytime. The style of the video is casual and seems to be a promotional or instructional video for cooking."
+    ],
+    "video_ids": [
+        "dHqwp1On8IQ_13_0to179"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chef in a white uniform, tray with three pieces of food, chef's smile? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chef in a white uniform holding a tray with three pieces of food, and the chef is smiling. These elements are clearly visible and match the description provided in the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chef in a white uniform holding a tray with three pieces of food. and the chef is smiling. The elements match consistent present and match the description provided. the question.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..63998f02c7e11d1faefac750b31591516d39f0b8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1337-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cee8e31509f82708e26c07c33a2df1aa250cb1acf957de699e6b3307f581ef09
+size 190458
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..e692cae958ddfb270395c5549f045ad976be4777
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man taking a selfie in a public square. He is wearing a black t-shirt and a backpack. The square is bustling with people, and there are benches and umbrellas scattered around. The man is smiling and looking directly at the camera. The background features tall buildings and a clear blue sky. The style of the video is casual and candid, capturing a moment of the man's day in a lively urban setting."
+    ],
+    "video_ids": [
+        "TvfePq94NUg_133_0to190"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a black t-shirt, a backpack? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a black t-shirt and carrying a backpack. These elements are consistent with the 'Object(s)' condition and are prominently visible throughout the frames.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man wearing a black t-shirt and a a backpack. The elements match consistent with the 'Object(s)' condition provided there the featured in the frames.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..15641e660ce691f437b4c3bf6d95d5396394d012
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:757a7c593e2cbd2226faac1289b9bc77c55a15ea351ed93958b4ce252c74b809
+size 286174
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..e9fca3cea671c703a345005988050715db6c4721
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a DIY aquarium setup on a wooden table. The scene features a clear plastic bottle repurposed as a water filter system connected to a rectangular glass tank filled with colorful gravel, artificial plants, and a small goldfish swimming gracefully among them. A hand enters the frame, adjusting the white tube attached to the bottle's cap, which is part of the filtration mechanism. The hand ensures the tube is properly positioned, emphasizing the importance of maintaining the setup for the fish's well-being. The background remains consistent throughout, focusing solely on the aquarium and the hand's interaction with the filtration system. The video captures the simplicity and ingenuity of creating a functional aquarium using everyday materials, highlighting the care taken in setting up an environment suitable for aquatic life."
+    ],
+    "video_ids": [
+        "4e921f56e64504f90d94f8fd7a1e99897ae07a1056df4601b39ab4984a18f7aa"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Rectangular glass tank, clear plastic bottle, colorful gravel, artificial plants, small goldfish, hand, white tube? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a rectangular glass tank, a clear plastic bottle, colorful gravel, artificial plants, small goldfish, a hand, and a white tube. All these elements are present and consistent with the description, with no contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as showcasing showing a rectangular glass tank with a clear plastic bottle, colorful gravel, artificial plants, a goldfish, a hand, and a white tube. The these elements are present and match with the description. and the additional or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9ada3a8589837cc006a23b5f9b94309db83efd7b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37dafca14a242cd601c0bf3a4948c368f7b4d3fc61dae4418ef6c46d48f64abb
+size 125835
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..846763a675dc8194b2c000f9f33ee052e9962753
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a group of people are gathered in a wooded area, where a man is seen cutting a log with an axe. The man is wearing a blue jacket and orange pants, and he is using a large axe to chop the log. The log is placed on the ground, and the man is standing over it, focusing on his task. The other people in the group are standing around, watching the man as he works. The setting is a forest, with trees and bushes surrounding the group. The atmosphere is calm and peaceful, with the sound of the axe cutting through the log being the main focus. The video captures a moment of outdoor activity, with the man demonstrating his skill in wood-cutting."
+    ],
+    "video_ids": [
+        "52HFgW6LmOQ_2_0to128"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a log, and a group of people? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man actively chopping a log with an axe, and there is a group of people in the background observing or participating in the activity. The core elements described \u2014 a man, a log, and a group of people \u2014 are all present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man in chopping a log with an axe. which there are a group of people observing the background observing. standing. the activity. The presence elements of in a man, a log, and a group of people \u2014 are all present and fulfilled depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8364671cad3b752167eac09e5fb4746c5abbed0a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e13a6f928b037e2960082545133a5faa870d763dcca66dd3a2766de90329eccf
+size 306180
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..0cbce5faf06ced97948cea6aa283ad0e1accbb6a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a car's steering wheel and dashboard, focusing on the details of the vehicle's interior. The steering wheel is black with a silver emblem in the center, and the dashboard features a digital display with various gauges and indicators. The car's interior is sleek and modern, with a focus on functionality and design. The video is likely a promotional or informational piece, showcasing the car's features and design elements. The style of the video is straightforward and clear, with a focus on the car's interior and the details of the steering wheel and dashboard."
+    ],
+    "video_ids": [
+        "L75e8kehniQ_31_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, digital display with gauges and indicators? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a steering wheel with the 'GT-R' logo prominently displayed on the center, which matches the 'Steering wheel' condition. Additionally, the digital display with gauges and indicators is visible on the left side of the steering wheel, fulfilling that part of the description. The presence of other interior elements does not contradict these core objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a steering wheel with a logoM'' logo in displayed in the center. which is the descriptionSteering wheel' part. Additionally, the digital display in variousuges and indicators is visible, the steering side of the steering wheel, fulfilling the part of the '. There video of the elements elements like not contradict the core elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..03e5d7f635850f8bf001f75d04625fff9802de71
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3ecf74b138ad33550188dd95cf164e8d95add75267efbcd8c6e76bdf72fea72
+size 154355
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..1260a79932eb5a06805b2312ea989426060ecbee
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a stylized, artistic representation of a young boy's fascination with the moon. In the first frame, the boy is seen looking up at the moon, which is depicted as a large, glowing orb in the sky. In the second frame, the boy reaches out towards the moon, as if trying to touch it. In the third frame, the boy's hand makes contact with the moon, and the moon begins to shrink, eventually disappearing into the boy's hand. The style of the video is dreamlike and surreal, with a focus on the interplay between the boy and the moon. The use of light and shadow creates a sense of depth and dimension, while the close-up shots of the boy's face and hand add a sense of intimacy and connection to the viewer."
+    ],
+    "video_ids": [
+        "BNGIpkROUrg_2_0to119"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Young boy, moon? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a young boy in profile, gazing at a large, detailed moon that appears to be floating or held in his hand. The core elements described \u2014 a young boy and the moon \u2014 are both present and accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a young boy interacting two, gazing at a large, detailed moon. dominates to be in in projected in his hand. The moon elements of\u2014 a young boy and a moon \u2014 are present present and central depicted in any.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..21fe1ef5c5e098cdb37cc81b5d8e916b9cd97c3c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f1ea025831ee1723f9f25e353d4c265c17a761f1675cec1b75b0c07c29921dd
+size 193079
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..ac1b711bfb4479f0c202a9963ed275791d3620f6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of camaraderie and celebration between two football players. The player in the foreground, wearing a white jersey with the number 27, is pointing towards the sky with his right hand, a gesture often associated with acknowledging the crowd or celebrating a successful play. His teammate, wearing a white jersey with the number 11, is standing next to him, smiling broadly and looking towards the camera, his left hand raised in a fist as if he's cheering or celebrating. Both players are wearing helmets with the Jaguars logo, indicating they are part of the Jacksonville Jaguars team. The background is filled with a crowd of spectators, their faces a blur of anticipation and excitement. The style of the video is dynamic and energetic, capturing the spirit of the game and the enthusiasm of the players and the crowd."
+    ],
+    "video_ids": [
+        "o2qlntM3qpo_10_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two football players in white jerseys with numbers 27 and 11, both wearing helmets with the Jaguars logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two football players wearing white jerseys with numbers 27 and 11, both wearing helmets with the Jaguars logo. The players are clearly identifiable as Jacksonville Jaguars based on their uniforms and helmets, and there are no elements that contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows two football players wearing white jerseys with the 27 and 11, both wearing helmets. the Jaguars logo. The players are in visible by wearing Jaguars players on the uniforms and helmets. which the are no additional in contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cad0b8ab8c3db1a24f25c80b8103bcd830c8c072
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ebecf3ae81d304a8ea176de5677a32a1e63fc64fdf4cbd7af8f764335a63595
+size 303199
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..d09da05a9cfe539dbebf8982cff5ae52e26e6c90
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a black and white camouflaged sports car driving down a road. The car is sleek and low to the ground, with large black wheels and a distinctive rear wing. The camouflage pattern is intricate and covers the entire body of the car, making it difficult to discern any specific design features. The road is surrounded by lush green trees and a fence, suggesting a rural or semi-rural setting. The car is moving at a moderate speed, and the driver is visible through the windshield. The overall style of the video is dynamic and action-oriented, with a focus on the car's design and movement."
+    ],
+    "video_ids": [
+        "NcchCD14HJk_8_627to777"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black and white camouflaged sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a sports car covered in black and white camouflage, which matches the description. The car's shape and design are consistent with a high-performance sports car, and the camouflage pattern is clearly visible throughout the video. No elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a black car with in a and white camouflage patterns which matches the description of The car's design and design are consistent with a sports-performance sports car, and the camouflage pattern is clearly visible. the video. The additional in the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f0cf9f9f40db59601ea1b0bbf3182a7718df9d37
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34427c3fd4837ae003079defd027cacd8aee6e8879627866a8766a927f239d45
+size 330404
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..bb2ad0ba896fc170924793e60fa66eaad8feed84
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene sunset scene featuring a large tree silhouetted against a vibrant orange sky. The sun is positioned low on the horizon, casting a warm glow that illuminates the branches and leaves of the tree. Perched on the branches are several birds, their dark forms contrasting sharply with the bright background. As the video progresses, one bird takes flight, adding a dynamic element to the otherwise tranquil setting. The background reveals a series of rolling hills or mountains, their outlines softened by the distance and the warm light of the sunset. The overall atmosphere is peaceful and picturesque, emphasizing the natural beauty of the moment."
+    ],
+    "video_ids": [
+        "06f2bd1ea4941d45781c5844023aad0213054af156ff91fe4f5643e00282dd27"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large tree, several birds perched on its branches, and rolling hills or mountains in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a large tree with silhouetted branches and leaves, several birds perched on the branches, and rolling hills or mountains in the background, all set against a sunset. The core elements described are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts depicts a large tree with ahouetted branches and leaves, which birds perched on its branches, and rolling hills or mountains in the background. which of against a sunset sky The presence elements of in present represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c42475fedc62488c37d57fc7a726f0b0781b3d4a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1338-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6747f813b549afffe9fe3a7790b80395be943c3538c92e20d7a2e431e703afca
+size 223574
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..dad9de5f4bd8b014c058be0ca3c15249077c7e93
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene lakeside scene, where the calm water reflects the lush greenery of the surrounding forest. The sky above is a clear blue, dotted with fluffy white clouds, adding to the tranquil atmosphere. The lake's surface is mostly still, with gentle ripples occasionally disturbing the reflection of the trees and sky. As the video progresses, a school of fish becomes visible near the bottom of the frame, swimming gracefully through the water. The camera remains stationary throughout, focusing on the peaceful interaction between the natural elements. The red structure on the left edge of the frame adds a pop of color against the predominantly green and blue hues of the landscape. The overall ambiance is one of quiet beauty and natural harmony."
+    ],
+    "video_ids": [
+        "c1553c73c3fe8fff5e39837190cadb74972c36a7be6ac8ba509cb561c495418f"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: School of fish, red structure? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a school of fish swimming in the lake, and a red structure is visible on the left side of the frame. These elements are consistent with the 'Object(s)' condition described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows shows a school of fish swimming in the water, which there red structure is visible in the left side of the image, The elements match consistent with the 'Object(s)' condition provided.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..92669a695cf7e9a5c00e8a0be502ba160752cffb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cde39bbce1d7c97cb326a90d6ed32dc8e23be0fdc48ec6e9b9580ddd044c1964
+size 185648
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..c7b52c7efa959cc9fa82fa06a4c69ffe4225e65e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a whimsical scene with a cartoon character resembling a Christmas elf. The elf, dressed in a red and white suit with a green hat, is seen in three different positions. In the first frame, the elf is standing upright on a yellow platform. In the second frame, the elf is leaning against a blue and red block. In the third frame, the elf is lying down on the yellow platform, seemingly asleep. The background consists of a white wall and a wooden archway, adding to the playful and imaginative atmosphere of the video. The style of the video is cartoonish and colorful, with a focus on the elf character and its interactions with the colorful blocks."
+    ],
+    "video_ids": [
+        "7yRFRPayfc4_68_0to150"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A cartoon character resembling a Christmas elf, a yellow platform, and blue and red blocks.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a cartoon character resembling a Christmas elf, wearing a green hat with a red feather and a red outfit. The character is sitting on a yellow platform, which is part of a structure made of blue and red blocks. These elements are clearly visible and match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a cartoon character that a Christmas elf, which a green hat and a red and, a red and with The character is standing on a yellow platform. which is consistent of the larger resembling of blue and red blocks. The elements match consistent visible and match the description provided.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fc214b8e4b8154b4057facf4da75e553f3077de1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:653bf0b3d787337e6329d0957570405b30c12b0b3df93caacd3a04c3c2d43c32
+size 219884
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..aaad967b890c7169affbf8800c8d210e5a533a68
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a bird in flight over a calm body of water. The bird, likely a duck or similar waterfowl, is seen mid-flight, its wings spread wide against the tranquil backdrop. The water below reflects the surrounding greenery, creating a mirror-like effect that enhances the peaceful atmosphere. As the bird glides gracefully across the surface, its reflection dances on the water's surface, adding a dynamic element to the otherwise still scene. The camera follows the bird's movement, maintaining a steady focus on the subject while subtly panning to keep it centered in the frame. The overall ambiance is one of tranquility and natural beauty, with the bird's flight providing a sense of motion and life amidst the stillness of the water and reflections."
+    ],
+    "video_ids": [
+        "4c21a68824dd8ececaec060b7c140bb42e4c27e7088fba370f4adb0be8b9eed7"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bird (likely a duck) in mid-flight.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bird in mid-flight over a body of water, with its reflection visible below. Although the bird's appearance is not clearly that of a duck (it has a more slender, streamlined shape typical of a waterbird like a heron or cormorant), it is indeed a bird in flight, fulfilling the core condition of the prompt.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a bird, the-flight, a body of water, which its wings visible on. The the bird's wings is somewhat entirely detailed of a duck,it could a more slender and elong body), of a her bird like a heron or aormorant), the fits in in bird in flight. which the ' condition of the '.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7adde75b537259ef6cddf99868096a7d5507ac18
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85e01bff5692c0ad35a308ef47a2839b4898d600548d1befe3212cac6f219aba
+size 172003
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..aa3304235cd88273e37a0e1445f7f17ce9620ea7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two men are seen in a dark setting, possibly a parking lot or a street at night. The man on the left is wearing a black jacket and a baseball cap, while the man on the right is wearing a plaid shirt and glasses. They are both holding skateboards, which are covered in colorful stickers. The skateboards are held horizontally, and the men appear to be examining them closely. The overall style of the video is casual and informal, with a focus on the skateboards and the interaction between the two men."
+    ],
+    "video_ids": [
+        "c0DwNTjtRMA_34_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, two skateboards? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men standing together, and both are holding skateboards that are covered in stickers. The core elements of the description \u2014 two men and two skateboards \u2014 are accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two men each side, each each are holding skateboards. are distinct in colorful. The presence description of the description, two men and two skateboards \u2014 are present represented in the video.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..36e53151d6184732a661c3a86c52e86b9d7e757e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eae9f902943948379c392aabd27a1c50a826dd79645236d0943758187ccb734f
+size 276477
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..37dc419938f8adb218ce63cd4053e109ea46c258
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures an intense moment in a football game. Two players, one in a white jersey and the other in a red jersey, are both leaping into the air to catch a football. The player in the white jersey is slightly ahead, reaching out to grab the ball with his right hand. The player in the red jersey is close behind, also reaching out with his right hand. The crowd in the background watches the action unfold, their attention focused on the two players. The video is shot from a low angle, emphasizing the height of the players' jump and the intensity of the moment. The style of the video is dynamic and action-packed, capturing the excitement and competition of the game."
+    ],
+    "video_ids": [
+        "wCL2a1Ty7so_6_0to160"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two players, one in a white jersey and one in a red jersey, and a football.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two football players in mid-air, one wearing a white jersey (number 8) and the other in a red jersey (number 26), both reaching for a football. The core description of the two players and the football is accurately represented, even though the red jersey is more maroon than pure red, which is a minor deviation and does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts two players players, action-air, one wearing a white jersey andnumber 3) and the other in a red jersey (number 93). both attempting for a football. The presence elements of two ' players and the football is accurately represented in and though the jersey jersey number not accuratelyoon than red red and and is a minor detail from does not significantly the description description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c691e1a2e5fe822d83d607e69ce248bef06f5bce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ca577bf047a0fdb88a9fbd6f7385d29bc8220326e7d476396316655f85ca2ba
+size 304892
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..d7e45bd0c21e5c0f975d2f15cd8a754851404c85
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with glasses, who appears to be a trial manager for a marine management organization. He is seated in front of a computer monitor, which displays a video of a boat. The man is wearing a striped shirt and is looking directly at the camera, suggesting that he is being interviewed or giving a presentation. The style of the video is professional and informative, with a focus on the man's expertise in marine management. The background is minimalistic, with the computer monitor being the main point of interest. The overall tone of the video is serious and focused, with the man's demeanor conveying confidence and knowledge in his field."
+    ],
+    "video_ids": [
+        "Qj8nOTwEzoA_22_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with glasses wearing a striped shirt, a computer monitor displaying a video of a boat.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with glasses wearing a striped shirt, and behind him, there is a computer monitor displaying what appears to be a video of a boat or maritime equipment. The core elements described are present, and no significant contradictions are observed.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing glasses wearing a striped shirt, and in him, there is a computer monitor displaying a appears to be a video of a boat. a scene. The description elements of in present, and there additional contradictions are noted.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f04f14f52acbe3e1c655336686d32ead2fad3f65
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:019d9d79f5394c88277fca701912ae6d3a71cd7b75fbcedaeab3aa1aa98c72cf
+size 207027
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..8ebdf03d24ffec75e0406b983ca8698befa5dc20
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a cooking tutorial featuring a dish made with green noodles, tomatoes, and avocado. The style of the video is casual and homey, with a focus on the food preparation process. The chef, Veronica Bosgraaf, is not visible in the video, but her presence is implied through the use of her book as a backdrop. The table is set with two plates of the finished dish, a bowl of pesto, a jar of olive oil, and a variety of fresh ingredients including tomatoes, avocado, and nuts. The chef uses a wooden spoon to mix the ingredients, and a fork to serve the dish. The overall atmosphere of the video is inviting and cozy, with a focus on the simple pleasure of cooking and enjoying a delicious meal."
+    ],
+    "video_ids": [
+        "63zyJ2H6QmY_3_127to312"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two plates of finished dish, a bowl of pesto, a jar of olive oil, tomatoes, avocado, nuts, a wooden spoon, and a fork.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing two plates of finished dish (pasta with pesto), a bowl of pesto, a jar of olive oil, tomatoes, avocado, nuts, and a fork. Although a wooden spoon is not explicitly visible, the utensils present are consistent with the described items, and the core elements are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showing showing two plates of finished dish,greenasta with vegetableso and a bowl of pesto, a jar of olive oil, tomatoes, avocado, and, a a wooden. The the wooden spoon is not visible mentioned, it presenceil in in consistent with the context objects. and the presence elements of accurately represented.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..69430875d75268ef5134661dde6011157de5db7d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d2fa10da361a3654e5c9eda4578465fbbd11e6d8888023dfcc4cdf8b0a498375
+size 217234
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..38be1ee5705e5f4e9130e9efa86474d4b0bbc6c1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up shot of a tree with a variety of fruits growing on it. The fruits are in different stages of ripeness, with some being green and others yellow. The tree is lush and full, with a dense canopy of leaves. The fruits are hanging from the branches, and the tree appears to be healthy and thriving. The video is likely educational, showcasing the growth and ripening process of these fruits. The style of the video is straightforward and informative, focusing on the natural beauty and growth of the tree and its fruits."
+    ],
+    "video_ids": [
+        "mAtw7rN7fco_48_22to172"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A variety of fruits growing on the tree, in different stages of ripeness (green and yellow).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows papayas on a tree in various stages of ripeness: some are green (unripe), while one prominently displayed fruit is yellow (ripe). This matches the condition of a variety of fruits growing on the tree, in different stages of ripeness.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts aayas growing a tree, various stages of ripeness, some are green,unripe), some others is displays is is yellow (ripe). This matches the description of having variety of fruits growing on the tree, in different stages of ripeness.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d2b412e0deeebd58fc44349ce33cabd49f52762f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1339-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6a4bea3d48f59c8ac00381ff8e503a1807b72bb3cd478ffe6efe2634dca7f0c6
+size 77323
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..48c59dd84c246a1864f6e094490ac6e1d4666f7c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a child's room with a white and teal bookshelf. The bookshelf has a white top and is filled with children's books. A small potted plant is placed on the top of the bookshelf. The room has a white wall and a wooden floor. The bookshelf is positioned against the wall. The video captures the simplicity and child-friendly environment of the room."
+    ],
+    "video_ids": [
+        "BcY-sDQB9c0_59_0to167"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white and teal bookshelf, children's books, a small potted plant? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white and teal bookshelf with children's books neatly arranged on its shelves. A small potted plant is also visible on top of the bookshelf, matching the description. Additional elements like toy blocks and a wall outlet do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white and teal bookshelf with multiple's books on arranged on the shelves. There small potted plant is also visible on the of the bookshelf. fulfilling the description provided The elements such the cars and a small-mounted are not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1e8ff9769d90e5cf578a03d064ecc621443269d3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2cf1bd126ed08ca8afee17d241abd9356e9afa7f48376f841940ccaa71b2fed6
+size 67343
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..99029369afda4e04d0319325e3a4548d477c3f21
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a football player from the San Francisco 49ers, specifically from the secondary team. The player is seen in three different frames, each capturing a different moment of the game. In the first frame, the player is seen with a focused expression, holding a football in his mouth, ready for action. In the second frame, the player is seen in a defensive stance, eyes locked on the ball, ready to intercept. In the third frame, the player is seen celebrating a successful play, his arms raised in triumph. The video captures the intensity and excitement of the game, showcasing the player's skill and dedication to the sport."
+    ],
+    "video_ids": [
+        "l6_GXmJ7VU8_97_0to138"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A football player from the San Francisco 49ers' secondary team.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a San Francisco 49ers jersey, with the text overlay explicitly stating 'SAN FRANCISCO 49ERS THE SECONDARY'. The player's attire, the team branding, and the context provided by the text all align with the description of a player from the 49ers' secondary team. There are no conflicting elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a football player wearing a yellow Francisco 49ers uniform, which the number ' ' stating '4 FRANCISCO 49ERS'. SECONDARY TEAM The player is uniform and including jersey's, and the context of by the text all align with the description of a player from the San49ers' secondary team. The are no elements elements in would this description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6fe94ca9eb6d4ba53513f9562c20e43b5ba72bff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:826b4e031fc3c211548356273e978fd13106efb22e9c7b5e1a1f3bc4de950d72
+size 334309
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..2f43657819062bec8c993567f530853ba1bcdce9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a cooking show featuring two women in a kitchen. They are both wearing red aprons with the words \"Cool Country\" printed on them. The kitchen is well-equipped with a stove, oven, microwave, and various cooking utensils. The women are preparing a meal, with one woman stirring a pot on the stove and the other woman placing food on a baking sheet. They are both smiling and seem to be enjoying the cooking process. The overall style of the video is casual and friendly, with a focus on the cooking process and the interaction between the two women."
+    ],
+    "video_ids": [
+        "ELS3mP9pa1o_46_27to237"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two women, a pot on the stove, a baking sheet, various cooking utensils.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two women in a kitchen setting, both wearing aprons. There is a pot on the stove, and a baking sheet with ribs is visible on the stovetop. Although not all cooking utensils are explicitly shown, the presence of the pot, baking sheet, and the context of cooking largely fulfills the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two women in a kitchen setting. with wearing redons with One is a pot on the stove, and a baking sheet with what on visible. the counterovetop. Various the all cooking utensils are explicitly mentioned, the presence of the pot and baking sheet, and the women of cooking suggest fulfillfills the 'Object(s)' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7d20f7f5ad7ee621801d9424a883037012ece75b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd5861470a54c735dc31034ee04e465347618b01bbb42d0bfeec368d49930d28
+size 196617
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..df61f41a2b7d03903004aa0cff7abdbd69d1a893
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a blue car with a unique design, featuring a pattern of white squares on its hood. The car is parked inside a garage, which is filled with various tools and equipment, indicating that it might be a workshop or a repair shop. The lighting in the garage is bright, with multiple fluorescent lights illuminating the space. The car is the main focus of the video, and its design suggests that it might be a custom or modified vehicle. The garage appears to be well-equipped for vehicle maintenance and repair, with shelves and cabinets filled with tools and parts. The overall style of the video is straightforward and documentary, with no additional narrative or embellishment. The focus is solely on the car and its surroundings, providing a clear and detailed view of the vehicle and the environment in which it is being showcased."
+    ],
+    "video_ids": [
+        "FyvoHkpZ6So_28_0to166"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue car with a unique design featuring a pattern of white squares on its hood.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a blue car with a distinct pattern of white squares on its hood, which matches the description. The camera pans across the hood, clearly displaying the pattern, and the surrounding garage environment does not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a blue car with a distinctive pattern of white squares on its hood, which matches the description provided The car angle around the car, clearly displaying the pattern. and the car environment environment does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..87ef14a45aafa6b1614e043481c586d1c86e40a7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ca75397528ebe4a93cecc2540c23d2e155057565f94b6652fa3c49cd13f04d8e
+size 171896
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..477f4a14065d4532d423a8d24be6cdf21936e39e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a gray sports car parked on a driveway. The car is sleek and modern, with a low profile and a streamlined design. The car is parked in front of a house, and there is a hedge and a tree in the background. The car is stationary, and there are no people or other vehicles in the frame. The style of the video is straightforward and unembellished, focusing solely on the car and its immediate surroundings. The lighting is natural, suggesting that the video was taken during the day. The car is the main subject of the video, and there are no other significant elements or actions depicted."
+    ],
+    "video_ids": [
+        "OMPD4d2YaWM_31_17to157"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A gray sports car with a sleek, modern design.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a gray sports car with a sleek, modern design, consistent with the description. The car's aerodynamic shape, sharp lines, and overall appearance match the characteristics of a modern sports car. The color is gray, and there are no elements in the video that contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a gray sports car with a sleek, modern design, which with the description. The car's aerodynamic shape, smooth lines, and the aesthetic align the characteristics of a sports sports car. The setting and a, and the are no elements in the video that contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e03be59e4a2e34db5295b0ca11b3594968bb6f8e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c756ed10c9b28bce9491d9ad44eb40cf7e2104fdb1ae80512e143c6752d114c
+size 259301
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..f8533ae8691993eaf7d4fd9414cb7034f10d1c71
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a football player in action, wearing a red helmet and a red jersey. The player is seen in three different frames, each showing a different emotion and action. In the first frame, the player is seen with a determined look on his face, his mouth open as if he's shouting or calling out to his teammates. In the second frame, the player's expression changes to one of triumph, his mouth still open but now with a smile, indicating a successful play or a victory. In the third frame, the player's expression changes again, this time to one of frustration or disappointment, his mouth still open but now with a frown, suggesting a failed play or a loss. The player's helmet and jersey are clearly visible in all three frames, emphasizing the intensity of the game and the player's emotions. The video is a dynamic portrayal of the highs and lows of football, capturing the player's emotions and actions in real-time."
+    ],
+    "video_ids": [
+        "e69MPU8jnf0_33_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A football player? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person wearing a football helmet and uniform, consistent with a football player. The details such as the helmet, gloves, and jersey align with the description of a football player, even though the image is static and appears to be a still frame or a very slow-motion capture.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a person wearing a red helmet and a, which with the football player. The helmet such as the helmet design uniform, and the are with the description of a football player.\"\n even though the video is AI and does to be AI close from from a very short-motion sequence.}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..82322f29e27d02cd1f3d45c6328c2843a4035111
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3c79b8f4f66ed72ded9c44b23d063759556abfada94a360f89a917ed38dc03d
+size 184736
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..bc9b3cdd24a25523d3c6fb323a79d16f54fd6826
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment from CTV News Channel, featuring two individuals in a studio setting. The first individual is a man with gray hair, wearing glasses, a suit, and a tie. He is seated in front of a backdrop that shows a cityscape with a body of water. The second individual is a woman with short hair, wearing a dark dress with a light-colored top. She is seated in front of a backdrop that shows a studio with multiple monitors and a news ticker. Both individuals are engaged in a conversation, likely discussing a topic of current interest. The style of the video is professional and informative, typical of a news broadcast."
+    ],
+    "video_ids": [
+        "AL8YG3LdS2M_10_26to211"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two individuals - a man with gray hair and glasses, wearing a suit and tie; a woman with short hair, wearing a dark dress with a light-colored top.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two individuals as described: a man with gray hair and glasses, wearing a suit and tie, and a woman with short hair wearing a dark dress. The woman's attire is consistent with a dark dress, and while the description mentions a 'light-colored top,' it is likely referring to a light-colored undershirt or collar, which is not clearly visible. The core description is largely accurate.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two individuals, described: a man with gray hair and glasses wearing wearing a suit and tie, and a woman with short hair, a dark dress with The setting's top includes a with a dark dress with but the the description does a lightlight-colored top,' it is not that to the light-colored blousehirt or blouse, which is not clearly visible in The presence elements of largely fulfilled,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1206341334220ade45ef4aa8fa6e6196e3d54a54
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6ad9240d021860c3c7a7dd2c430e993ae85dc34f6320a7c9ce51e35280d6f279
+size 143174
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..46335abc55e9a08484dbea0eedd345d4afee02b6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the festive spirit of Christmas with a beautifully decorated tree. The tree is adorned with gold ornaments and twinkling lights, creating a warm and inviting atmosphere. The tree is set against a backdrop of wrapped gifts, adding to the anticipation of the holiday season. The camera pans around the tree, showcasing its full glory and the meticulous arrangement of the ornaments. The overall style of the video is festive and joyful, capturing the essence of Christmas celebrations."
+    ],
+    "video_ids": [
+        "SVZ9Bt6gF-U_2_0to198"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Christmas tree adorned with gold ornaments and twinkling lights, wrapped gifts? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a Christmas tree decorated with gold ornaments and twinkling lights, surrounded by wrapped gifts. These elements directly match the described 'Object(s)' condition without contradiction, even though additional decorative elements like red plaid wrapping paper are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a Christmas tree decorated with numerous ornaments and twinkling lights, which by wrapped gifts. The elements align match the description 'Object(s)' condition, any.\"\n fulfilling though the elements elements like rib ribaid wrapping are are present,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a492529b84550ee4adfcdf5b0ce86094fe76d121
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-134-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e3d9776dc43ad7671f2b2656778abfb1684c24ee82e48120521a6010c0621b4
+size 234276
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..1a7180efe38cb22cfc5646c2d06a5f317f106d65
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man and a woman sitting on a couch, engaged in a conversation. The man is wearing a plaid shirt, while the woman is dressed in a red sweater. They are seated on a blue couch, which is positioned in front of a window. The window offers a view of a cityscape, with buildings and boats visible in the background. The overall style of the video suggests a casual, relaxed atmosphere, with the focus on the interaction between the two individuals."
+    ],
+    "video_ids": [
+        "01U5dZOkEv4_6_41to208"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman sitting on a couch, with the man wearing a plaid shirt and the woman in a red sweater.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a woman sitting on a couch. The man is wearing a plaid shirt, and the woman is wearing a red sweater. These elements match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a woman sitting on a couch. The man is wearing a plaid shirt, and the woman is wearing a red sweater. The elements match the description provided. the questionObject(s)' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e1f2540748a91de862a9467966a4402bf35cafb0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:363a7ee831aef3b48dcc1caa3c4ba413115ea84424148d1322f60c46e8a02379
+size 184534
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..e63bee5a5dee64c8d75d22a4d17b3164bcc2caec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a rusted engine, likely from a motorcycle or car, set against a grassy field with a small pond in the background. The engine is dirty and appears to be in disrepair, with visible rust and missing parts. The style of the video is realistic and documentary, focusing on the details of the engine and its surroundings. The camera angle is low, emphasizing the size and complexity of the engine. The lighting is natural, with the sun casting shadows on the engine and the grass. The overall mood of the video is somber and contemplative, as it captures the passage of time and the effects of weathering on once-functional machinery."
+    ],
+    "video_ids": [
+        "bbCVoRPv96w_26_0to122"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Rusted engine? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a motorcycle engine that appears heavily rusted, with visible corrosion on its metal surfaces. This matches the 'Rusted engine' condition described. The background and other elements do not contradict this observation.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts a close-up of a motorcycle engine that appears to rusted, with visible signs and the surface surfaces. The align the descriptionRusted engine' condition as, The presence, other elements, not contradict this core.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..59b4d7ee32d0301dcfb442976a10565ba7513ed3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9de5d8484596af94e9748aa5f97d35d94a9c5ba06a29de41c871e6d1dda1f760
+size 225186
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..1a417a10165f49deaf1df0d6c70fd9d9ce647d03
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person standing in a room filled with motorcycle gear. The person is wearing gray pants with a zipper on the side and a black belt. They are also wearing a white and red jacket with a black and white helmet. The helmet is placed on a shelf in the background. The room has a white wall and a black floor. The person is standing in front of a window that has a white curtain. The window is letting in natural light. The person is looking at the camera. The video is in color and has a clear focus on the person and their gear."
+    ],
+    "video_ids": [
+        "WtcLaShdba0_18_22to203"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, gray pants with a zipper, black belt, white and red jacket, black and white helmet, motorcycle gear, window, white curtain? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by showing a person wearing gray pants with a zipper, a black belt, and a white and red jacket. In the background, there are black and white helmets and motorcycle gear visible. The setting appears to be indoors, possibly with a window and white curtain, although these are less prominent. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as depicting a person wearing gray pants with a zipper, a black belt, a a white and red jacket. The the background, there is black and white motorcycle, a gear,. Additionally window includes to be a with with in a window and a curtain, which the are not distinct in The elements elements of in present, contradiction.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..22cff8692eb690a4d8c36d71f82d6bccdd549f9b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a6bb8b85c9f045237b7e618fb8dc2f0684db84ee33c04d0a804a1fe0ba17eae2
+size 112253
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..0c2601634dd18e5d1e1f7ecf85a8cb1c96422bb2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person preparing food in a kitchen. In the first frame, the person is seen holding a pot filled with chopped potatoes. In the second frame, the person is seen adding more chopped potatoes to the pot. In the third frame, the person is seen stirring the contents of the pot. The kitchen counter is visible in all frames, and there is a snowman decoration in the background. The style of the video is a simple, everyday cooking scene."
+    ],
+    "video_ids": [
+        "qDbPJLiX8_Q_1_595to789"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person, pot, chopped potatoes? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person's hand interacting with a pot filled with chopped potatoes, fulfilling the 'Object(s)' condition. The presence of a snowman figurine and other kitchen items does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a person ( hands interacting with chopped pot filled with chopped potatoes. which the 'Person(s)' condition. The presence of additional snowman andine and other decorative items in not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ac131c8a01dff97217cb4a69ec392885f53d923b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6366e343b490495d46c2d311d84c6efa6d9b14cf9ec6e756b004611d51d9e1da
+size 219507
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..c041ccfe43221840b92a8a64bcf024e7c404a92e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows an elderly woman sitting in a chair, looking at a piece of paper. She is wearing a black and white patterned top and appears to be in deep thought or possibly experiencing a moment of sadness or confusion. The setting is a simple room with a wooden chair and a potted plant in the background. The lighting is soft and natural, suggesting an indoor environment. The style of the video is realistic and it captures a candid moment in the woman's life."
+    ],
+    "video_ids": [
+        "27_JskjGS9w_16_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An elderly woman sitting in a chair, looking at a piece of paper. She is wearing a black and white patterned top.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an elderly woman sitting in a chair, holding and looking at a piece of paper (a notebook), and she is wearing a black and white patterned top. These elements match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an elderly woman sitting in a chair, holding and looking at a piece of paper.object book). which she is wearing a black and white patterned top. The elements match the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c4bec37c29e92d3c5d7f89f79af80774d196a22b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:10af958ecf9aef01e5d8fbc2c4f54ceed29c84d2cc16f0478186d30380b7b468
+size 223619
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..008527f0b59c8049d3e3c4eb2ed532cec983b094
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a black vest standing in a room with a green wall. He is gesturing with his hands, possibly explaining something. Behind him, there is a shelf with various objects, including a chess set, books, and figurines. The room has a vintage feel, with a reel-to-reel tape recorder and a film projector on a table. The man appears to be in the middle of a conversation or presentation, as he is actively engaging with the viewer. The overall style of the video is casual and informative, with a focus on the man and his surroundings."
+    ],
+    "video_ids": [
+        "2wRcDM_5rjs_9_28to217"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black vest, a shelf with a chess set, books, and figurines, a reel-to-reel tape recorder, and a film projector.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black vest, standing in front of a shelf that contains a chess set, books, and figurines. A reel-to-reel tape recorder and a film projector are also visible on a table behind him. All specified objects are present and do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man in a black vest, standing in front of a shelf that contains a chess set, books, and figurines. There reel-to-reel tape recorder and a film projector are also visible on the table in him. The elements elements are present and match not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c586c3146c9bde876dc1705fbeb335e6e3b615a1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4a270ca072aa5d71dc5f76248ba1d1e3855b80df6ce20f21a32aef255ad06fc9
+size 124428
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..545981ad43f884490aa7376f098c80a8c74b9063
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a car's engine compartment, captured in three frames. The style of the video is a straightforward, unembellished documentation of the car's engine components. The engine is a complex assembly of various parts, including the air intake system, the fuel system, the cooling system, and the engine block. The colors of the engine components are predominantly metallic, with some black and gray hues. The engine appears to be well-maintained and clean. The video does not contain any text or additional graphics. The focus is solely on the car's engine, providing a detailed view of its internal components."
+    ],
+    "video_ids": [
+        "KcNvc_p5prU_0_0to135"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Air intake system, fuel system, cooling system, engine block? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows components that align with the 'Object(s)' condition: Air intake system (visible intake pipes and air filters), fuel system (fuel lines and injectors), cooling system (coolant reservoir and radiator components), and engine block (the main engine assembly). The visual details match these described systems.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful the that are with the 'Object(s)' condition, Air intake system,visible air manifold), air filter), fuel system (fuel lines and possiblyors), cooling system (cooling reservoir and radiator),), and engine block (the main structure structure). The presence elements in the systems systems,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5c08b6cfd950061842b045e434fdb5f97d521a0a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a2fbda2c16a25e50649187c3647dc829ff8f3f73daec4955e425af1c284982ca
+size 73734
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..d87a20457a48464044455e097172dd215cba5ca7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the process of frying chicken wings in a large, round, black frying pan placed on a red electric stove. The pan is filled with bubbling oil, indicating a high temperature suitable for frying. Several pieces of golden-brown chicken wings are sizzling in the oil, their surfaces glistening with the hot fat. The camera remains stationary throughout the sequence, focusing closely on the frying pan to highlight the cooking process. The oil bubbles vigorously around the chicken, suggesting that the wings are being cooked evenly. The scene is set in a kitchen environment, though the background is blurred, drawing attention solely to the frying action. There are no visible characters or additional objects in the frame, keeping the focus entirely on the frying process. The video emphasizes the transformation of the raw chicken into a crispy, appetizing dish through the continuous bubbling and sizzling of the oil."
+    ],
+    "video_ids": [
+        "89f0141ec868cf5d2c3d7cdc525fceb1c686cb2cd1940bfab7629e5ee9a8f37d"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Several pieces of golden-brown chicken wings? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows several pieces of chicken wings frying in oil, and they appear golden-brown, matching the description. The presence of a frying pan and oil does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows several pieces of food that that in a, which they appear to-brown, which the description of The presence of oil frying pan and the is not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dbce399fa19f057db251def8e09c8671d845e7c2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1340-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d1a61becc92dc75405c39cd52479698b6dd9e8a5001cfad68aa56ad2398e124d
+size 220286
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..16e81b1f4baf6ca52e9bb9d6192a77a61d6140cc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a fish being grilled over an open flame. The fish is positioned on a black, perforated grill surface, which is part of a larger grill setup. The grill is situated outdoors, as evidenced by the visible flames and embers beneath the fish. The fish itself appears to be a whole, raw specimen, likely a type of white fish given its pale color and texture. As the video progresses, the fish begins to cook, with slight changes in its appearance becoming noticeable. The skin starts to turn a golden brown, indicating the cooking process is underway. The camera remains steady throughout, focusing on the fish and the grill, providing a clear view of the grilling process. The background is slightly blurred, drawing attention to the main subject\u2014the fish and the grill. The overall scene suggests a casual outdoor cooking environment, possibly during a barbecue or a similar gathering."
+    ],
+    "video_ids": [
+        "998fae9ed2727430a0ad12b5416186fb0fd8059961fada1b050f83787747662a"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A whole, raw fish and a grill setup.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a raw fish placed on a grill setup with visible flames and coals underneath, matching the core description. The fish appears whole and uncooked, and the grill setup is clearly depicted with a metal grate over the fire, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a whole fish placed on a grill setup, a flames underneath emals,, which the description description of The presence appears whole and uncooked, and the grill setup is clearly visible with the grill grate and the fire. fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d562078229a7585291bee6df83b73c824a97fbaa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb2e3b38fc230fa0142026f47fd529578dffb6d454f1a1c8736dc31514e4145c
+size 156953
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..bf109d145720701dd80cbcc313358a7392bd3158
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a husky dog standing in a room with a white door. The dog is looking up at the camera with a playful expression. The dog's fur is fluffy and white, and it has a black and gray face. The dog is standing on a gray carpet. In the background, there is a purple ball on the floor. The dog is the main subject of the video, and it appears to be in a playful mood. The room is brightly lit, and the white door provides a contrast to the dog's fur. The purple ball adds a pop of color to the scene. The overall style of the video is casual and fun, with the dog as the main focus."
+    ],
+    "video_ids": [
+        "BlrF0DCDYH8_4_0to169"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A husky dog, a purple ball? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a husky dog standing indoors, and a purple ball is visible on the floor near the dog's front paws. These objects are present and consistent with the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a husky dog and in, and there purple ball is visible on the floor near the dog. front legsaws. The elements match consistent and match with the description provided.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..10ca02ff605ef7fdb3fede4fefc59bf2aaf5a89a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3da469607251d5fa5ab04bfff1d1dffe7dd42fe40d2725e734d52021ee490f0f
+size 166746
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..e6e8110e59da6763c41b0faa788af0df8f168eed
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a puffin swimming in calm, rippling water. The bird's distinctive black and white plumage is clearly visible, along with its bright orange beak and feet. The puffin is seen paddling vigorously with its wings, creating small splashes as it moves through the water. The background consists of gentle waves, suggesting a serene coastal environment. The BBC logo is visible in the top left corner, indicating the source of the footage. The camera remains steady throughout the sequence, focusing on the puffin as it swims from left to right across the frame. The overall scene conveys a sense of tranquility and natural beauty."
+    ],
+    "video_ids": [
+        "196033bd3cb251b74cf973ae87e5176877fe31835a0502c48a8102923225a975"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A puffin, gentle waves in the background, BBC logo in the top left corner.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a puffin swimming in gentle waves, which matches the description. The BBC logo is visible in the top left corner throughout the video. There are no elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a puffin swimming in water waves, which ful the ' of The BBC logo in visible in the top left corner, the video, The are no additional in contradict the given description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0b7361ac55169f47a1a107c74721596fd3c33a1d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5751d53697bbce3909e73ea0131d7ce3283e392fc8e659f80b3b939f1a3aca9
+size 228883
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..a8a2fc3b05a9ddf9bb84172bdccbc3c986d00cec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man sitting in a studio, speaking into a microphone. He is wearing a white hoodie and a jacket with a floral pattern. The studio has a cityscape in the background, and there is a logo on the wall that reads \"DIS\". The man appears to be engaged in a conversation, possibly a radio interview or a podcast recording. The overall style of the video is professional and polished, with a focus on the man and his interaction with the microphone."
+    ],
+    "video_ids": [
+        "7n3ETLy1r28_9_0to181"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a microphone, a white hoodie, a jacket with a floral pattern, a cityscape, a logo reading 'DIS'.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting in front of a microphone, wearing a white hoodie under a jacket with a floral pattern (roses). The background features a cityscape, and the logo 'DISH NATION' is visible, which includes 'DIS' as requested. All core elements of the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man wearing in front of a microphone, wearing a white hoodie and a jacket with a floral pattern.whiches and The background features a cityscape with and there logo 'DIS''ATION' is visible, which is theDIS'. as part. The elements elements of the description are present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..29a8c54eb6a98bb8fb2d51ffbaa1ffddafbaafb6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c0faea3335f91d26cfc24573c4f5413e96c00a27193e1dd42b14ae3a258fe6f5
+size 216311
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..3c6039978000e3b02a2d36e21b62de189ef18940
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a bustling city street in a foreign country. The street is lined with tall buildings, each adorned with colorful signs and awnings. The buildings are predominantly white and yellow, with some featuring balconies and shutters. The street itself is busy with traffic, including cars and a truck, all moving in the same direction. The vehicles are a mix of colors, adding to the vibrant atmosphere of the scene. The street is also populated with pedestrians, some of whom are carrying umbrellas, suggesting that it might be a rainy day. The overall style of the video is dynamic and colorful, capturing the energy and vibrancy of city life."
+    ],
+    "video_ids": [
+        "nYSNK0IJM4Q_41_0to133"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Tall buildings, cars, a truck, pedestrians? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing tall buildings in the background, multiple cars parked and moving along the street, a truck passing by, and pedestrians visible on the sidewalk. All specified elements are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing tall buildings on the background, multiple cars and and moving along the street, a truck, by, and pedestrians walking on the sidewalks. The these elements are present and contribute depicted,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1dc593fa97092e2d064893ff283c352cee619fb3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62f5bc195e74ee0eccb3090a1eb3e6ccf42d405fc5fa0ffc069d6235361d685a
+size 127173
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..562f00131f1aac0dcd3c8eb87255e78ab48cd2ce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man standing next to a black SUV in a parking lot. He is wearing a white t-shirt and a black baseball cap. In the first frame, he is looking at the camera with a slight smile. In the second frame, he is pointing at the camera with his right hand. In the third frame, he is giving a thumbs up sign with his right hand. The background of the video shows a clear sky and a building in the distance. The style of the video is casual and informal."
+    ],
+    "video_ids": [
+        "fUCNIuPCAnU_42_66to233"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a black SUV? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man standing next to a black SUV. The man is visible wearing a white t-shirt and a cap, and the SUV is prominently featured in the frame. There are no elements that contradict the core description of a man and a black SUV.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man standing in to a black SUV. The man is wearing in a white t-shirt and a cap, and the black is prominently displayed in the background. The are no additional in contradict the description description of a man and a black SUV.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..da8c432002247b2be10573b9ef66f339160840ef
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:65cfea72f4b09c4e36af854300999127b951f4699bf48c7f54f2cf9e3aeb2647
+size 179430
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..d2be0bc40973cb8c9d0828fe28056bc339bfca0e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car, focusing on the passenger seat. The seat is upholstered in a light beige or cream color, and it appears to be a luxury vehicle, given the high-quality materials and craftsmanship. The seat is empty, and the car door is open, revealing the door panel and window controls. The car's interior is well-lit, suggesting that the photo was taken during the day. The style of the video is straightforward and documentary, with no additional embellishments or effects. The focus is solely on the car's interior, providing a clear and detailed view of the passenger seat and door panel."
+    ],
+    "video_ids": [
+        "JqEKjuol_T0_32_0to156"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Passenger seat, car door, door panel, window controls? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the passenger seat, part of the car door (including the door panel), and window controls (implied by the presence of the door handle and window control switches on the door panel). The camera pans across these elements, providing a detailed view that aligns with the requested objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a interior seat, car of the car door,including the door panel and and window controls.asplied by the window of the window handle and window frame buttons). the door panel). The image angle to these elements, fulfilling a clear view of aligns with the description objects.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f0033f0e878df6bc4d1bfd35cae7353cda7137ae
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c0684382892d7ea117ff29ec8a4768234ff42eaf3cf33141c4c84eebcdfa36f
+size 117345
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..62e2f27f9bd4356b610bca007e862f4494838406
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and curly hair, standing in front of a white background with text. The text includes \"THE LINCOLN MOTOR COMPANY\", \"TRIBECA FILM FESTIVAL\", and \"WHAT\". The man is wearing a black jacket and a black shirt. The style of the video is a simple, straightforward portrait of the man, likely taken at an event or festival. The focus is on the man, with the text serving as a backdrop. The lighting is even, and the man is looking directly at the camera. The overall impression is of a professional, well-dressed individual at a formal event."
+    ],
+    "video_ids": [
+        "704EIYNFUnU_7_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and curly hair, wearing a black jacket and black shirt; text including 'THE LINCOLN MOTOR COMPANY', 'TRIBECA FILM FESTIVAL', and 'WHAT'.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and curly hair, wearing a black jacket and black shirt, standing in front of a backdrop that includes the text 'THE LINCOLN MOTOR COMPANY', 'TRIBECA FILM FESTIVAL', and 'WHAT'. All elements described in the condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and curly hair, wearing a black jacket and black shirt, which in front of a backdrop that includes the text 'THE LINCOLN MOTOR COMPANY', 'TRIBECA FILM FESTIVAL', and 'WHAT'. The elements of in the question are present in match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6e5cc9c0952cf8342619302a73130bf441320be1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1341-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f114b912b7a4c5d8158b3829a3e371f93cc44fb2b232cb16087f70e9fadb5e08
+size 135618
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..b12b5ac9ae01857428582c283b078f3085ce4d96
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man with a beard is standing in a kitchen, wearing a red and blue plaid shirt. He is looking directly at the camera with a surprised expression on his face. The kitchen is well-lit and features a sink, a window, and a potted plant. In the background, another person is visible, but they are not the focus of the video. The style of the video is casual and candid, capturing a moment of surprise or amusement in a domestic setting."
+    ],
+    "video_ids": [
+        "1a_drpzrZbE_30_0to151"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard wearing a red and blue plaid shirt, and another person in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man with a beard wearing a red and blue plaid shirt in the foreground, and another person is visible in the background, fulfilling the described 'Object(s)' condition. Additional elements like the kitchen setting and decorations do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man with a beard wearing a red and blue plaid shirt. the foreground. which another person is visible in the background. partially the ' conditionsObject(s)' condition.\"\n The elements like the kitchen setting and the do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0d51d1dbd34f32667e4dd925754a6fe673ab2d5f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8a05e63853216dddae78d5f30966a51901d34646e9e03bfa017ce026708eb5b9
+size 185035
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..5113afb0a086bb0b25f494075deac673d4c2acd7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man wearing a black t-shirt and green sunglasses. He is standing in front of a green hedge. The man is smiling and appears to be enjoying himself. The sunglasses are a bright green color and are quite large, covering a significant portion of his face. The man's t-shirt is black and he is wearing it with a casual style. The green hedge behind him provides a nice contrast to his outfit. The overall style of the video is casual and fun, with the man's bright green sunglasses adding a touch of color and personality to the scene."
+    ],
+    "video_ids": [
+        "7-arSliviBA_10_0to140"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man wearing a black t-shirt and green sunglasses.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black t-shirt and green sunglasses, which matches the description. The background consists of green foliage, but this does not contradict the core description of the man's attire and accessories.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black t-shirt and green sunglasses, which matches the description provided The background is of green foliage, which this does not contradict the core description of the man's attire. the.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f1fb79153159f9f2202c4d545dbe0a945a717a1c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c80d0cc4a5735276fde5856d3460aeeca6f6b0ff96e4fb499783fe300edc2f0
+size 186991
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..e0fdc5e20232a8d6ee46249a09c7ced401f29fd6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a variety of sliced bread and bagels arranged on a white plate. The bread and bagels are stacked in a pile, with some slices overlapping each other. The bread appears to be freshly baked, with a golden-brown crust and a soft, fluffy interior. The bagels have a dense, chewy texture and are covered in sesame seeds. The plate is placed on a white surface, which contrasts with the golden-brown color of the bread and bagels. The video is shot in a realistic style, with a focus on the textures and colors of the bread and bagels. The lighting is bright and even, highlighting the details of the bread and bagels. The video does not contain any text or additional elements."
+    ],
+    "video_ids": [
+        "Jveit7UKBy4_3_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Sliced bread and bagels arranged in a pile.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a pile of sliced bread and bagels arranged on a white plate, matching the described condition. The focus is on these items, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close of sliced bag and bagels, on a plate plate. which the description '. The bag is on the items, and there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b7bccafcdd8243905cbafd7b0f271a04dc95f2cc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:78e7ab06c213214b0fd31f4411c240714e174932b529f8609359e468795f2007
+size 233134
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..627d676b086d69488c3bd96f441fd96e55b0f6f8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man named Maxwell Thomas, who is an accordionist. He is standing in front of a microphone, ready to perform. The background is a stage with a blue and purple lighting, creating a vibrant atmosphere. The man is wearing a white tank top and has a beard, adding to his unique style. The video captures the anticipation of the performance, with the man looking directly at the camera, ready to entertain the audience. The overall style of the video suggests a live music performance, with the focus on the musician and his instrument."
+    ],
+    "video_ids": [
+        "9ta9dDz_OfI_2_42to218"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man named Maxwell Thomas, an accordionist, and a microphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man identified by on-screen text as \\\"MAXWELL THOMAS Accordionist\\\". He is holding an accordion and has a microphone positioned near him. The core elements described \u2014 Maxwell Thomas, an accordionist, and a microphone \u2014 are all present and correctly represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man holding as the-screen text as MaxwellMaxWELL THOMAS\\\"ordionist\\\" playing He is holding an accordion and there a microphone in in him, The setting elements of in a Thomas, an accordionist, and a microphone \u2014 are all present in match identified in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5ff19e418c56f4ed2b61f5e93acfc63a326d9cfc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37c68070cf1112fb8884f73830f6e512addf6bcb8ca463d71d6ed0e11c7aee67
+size 233900
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..da4b1de2de3d0fe7711d03ec09dc836a13346ce9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of intense emotion from a soccer player. The player, dressed in a white and blue striped jersey, is seen with his hand covering his mouth, a clear sign of shock or disbelief. The setting is a soccer stadium, with the blurred background suggesting a large crowd of spectators. The player's focused gaze and the intensity of the moment suggest that the game is at a critical point. The style of the video is realistic, capturing the raw emotion of the player in a real-life setting."
+    ],
+    "video_ids": [
+        "U1swfC0ZE3Q_9_0to103"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A soccer player dressed in a white and blue striped jersey, with his hand covering his mouth.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a white jersey with blue stripes, which matches the description. His hand is covering his mouth, and although the jersey has yellow accents, the core description of white and blue striped jersey with hand covering mouth is fulfilled. The presence of additional elements (like the logo and background) does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a white and with blue stripes, which matches the description. The hands is covering his mouth, which he the jersey has a and, it core description of the and blue stripes jersey is the covering the is fulfilled. The additional of the elements likethe the crowd and the) does not contradict the main description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..855f83b67a00645868c9f3800e582753b36aa722
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5508c60552e40135acf79030e2aad03d9b0fcd19b46ad921c802f453474adbd
+size 309801
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..c88986d4e7591c0308848920d068c975426ccdfd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with long, dark hair and a white sweater. She is standing in a room with a metallic wall and a wooden bowl on a shelf in the background. The woman is looking to her left with a slight smile on her face. The lighting in the room is soft and warm, creating a cozy atmosphere. The woman's expression and the overall setting suggest a casual and relaxed scene."
+    ],
+    "video_ids": [
+        "__18o0BCzNY_20_23to144"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with long, dark hair and a white sweater, a wooden bowl on a shelf.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a woman with long, dark hair and a white sweater, which matches the core description. In the background, there is a wooden bowl on a shelf, which is also present as described. No elements contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with long, dark hair wearing a white sweater, which matches the description description. Additionally the background, there is a wooden bowl on a shelf, which also also consistent in described. The additional contradict the given,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..360fa9640708107493fbee70f8f5a5f267197a93
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab5fad329c0568646c22c24d6bcbfc6c77679dfcd46abb09e559f95f68d158a2
+size 119189
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..a03d9d0e8c1a1af484a351974568a0c8bad66439
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a silver Mercedes-Benz convertible parked in a garage. The car is the main focus of the video, with its sleek design and shiny exterior. The garage is filled with various objects, including a black and white checkered airplane propeller, a red car, and a black car. The garage appears to be a workshop or storage area, with tools and equipment scattered around. The lighting in the garage is bright, highlighting the details of the car and the surrounding objects. The video is likely a promotional or showcase video for the car, highlighting its design and features."
+    ],
+    "video_ids": [
+        "6xt_sOgtHEs_11_0to126"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Silver Mercedes-Benz convertible, black and white checkered airplane propeller, red car, black car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a silver Mercedes-Benz convertible with its trunk open and door open, a black and white checkered airplane propeller (on a vintage aircraft in the background), a red car (visible in the background to the right), and a black car (visible in the background behind the convertible). All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a silver Mercedes-Benz convertible, a top open, a a, a black and white checkered airplane propeller,which the stand car), the background), a red car,part in the background), the left), and a black car (part in the background to the red). The elements elements of in present in contradiction.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1c36c8d766d5fe480e585178fa21e6392317c1f8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:072f62e163bfb1ec20c8b587a172918fad9317af7312a2a63aff34ecacf97d31
+size 156355
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..e5d841e9f5e5b30d34a87738b2c723c159dd2c6c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a luxury sports car, focusing on the steering wheel and dashboard. The steering wheel is black with a yellow Ferrari logo in the center, and it features several buttons and dials on the left side. The dashboard is also black, with a digital display showing the car's speed and other information. The car's interior is sleek and modern, with a focus on high-quality materials and attention to detail. The video is shot from a low angle, emphasizing the car's design and features. The style of the video is sleek and modern, with a focus on the car's interior and design."
+    ],
+    "video_ids": [
+        "NZdUAY1NwNU_31_0to104"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel (black with a yellow Ferrari logo, featuring several buttons and dials), Dashboard (black with a digital display showing the car's speed and other information)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a black steering wheel with a yellow Ferrari logo, along with multiple buttons and dials. The dashboard is visible in the background, appearing black with a digital display, although the specific details of the display (like speed) are not clearly visible. The core elements described are present and consistent with the video content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a steering steering wheel with a yellow Ferrari logo in which with several buttons and dials. The dashboard is visible in the background, displaying black with a digital display that which the specific details of the display arelike speed) are not clear discern. The elements elements of in present, match with the video content.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6160c6870901815457250c57048487939e1cc0e8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1342-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a3fefb1891a7bb505a833493aa1997e73a00acba601e7019145280787b6e57f
+size 132692
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..7fa5b17cd16a9c17f8b14d504b95d573eb7ef54d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man riding a motorcycle, with a focus on his facial expression and the motorcycle's details. The man is wearing a plaid shirt and a black tie, and he has a surprised or excited look on his face. The motorcycle has a blue and red flag attached to the front, and there is a certificate or award displayed on the back of the motorcycle. The background shows a cloudy sky and a building with a clock tower. The style of the video is dynamic and energetic, with a focus on the man's reaction and the motorcycle's details."
+    ],
+    "video_ids": [
+        "Y11ex5jM4ug_27_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a motorcycle, a plaid shirt, a black tie, a blue and red flag, a certificate or award.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man on a motorcycle wearing a plaid shirt and a black tie. There is a blue and red flag (Union Jack) draped over the motorcycle, and a framed certificate from 'Guinness World Records' is visible. All core elements from the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man wearing a motorcycle. a plaid shirt and a black tie. There is a blue and red flag visiblepossibly Jack) visible over the motorcycle, and the certificate certificate or theHondainness World Records' is visible on The the elements of the description are present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c4c7d59664b6c9a2d5443e439d6602b5892ae2c1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9048f28c2521a020d5e0dfdff7d6c6665028dbcb125c7e3c2e88f48dda14634b
+size 259516
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..197b32753015ae064e162f97365542ce0c5421e8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman with a short, curly brown hairstyle and red lipstick is seen in a room with a green curtain. She is wearing a blue top and is seated in a chair. The woman appears to be engaged in a conversation with someone off-camera, as she is looking to her left with a concerned expression on her face. The room has a warm and inviting atmosphere, with the green curtain adding a touch of color to the scene. The woman's attire and the setting suggest a casual, yet thoughtful interaction."
+    ],
+    "video_ids": [
+        "Bzc7VsqvjkI_10_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with a short, curly brown hairstyle and red lipstick, wearing a blue top, sitting in a chair.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with a short, curly brown hairstyle and red lipstick, wearing a blue top, sitting in a chair. These features match the description exactly, and no elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with a short, curly brown hairstyle and red lipstick, wearing a blue top, sitting in a chair. The elements match the description provided, indicating there additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b010828260dc6b5f444f319f27a299e7c9191d38
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:356b3f43bbaa136db99a500c8fac2aa75205efe0a1bba0663da3dff95c590283
+size 200973
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..a7d8b1e0ef80b97b4ef677459c031a580cd4d9d3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene winter scene on a snow-covered deck. The deck is blanketed in a thick layer of white snow, with footprints scattered across the surface, suggesting recent activity. At the center of the frame is a black plastic container, possibly a bird bath or a small water feature, which stands out against the white backdrop due to its dark color and reflective surface. Surrounding the container, the snow appears undisturbed except for the footprints.  In the background, a wooden structure with a railing is visible, providing a sense of enclosure and warmth amidst the cold environment. A staircase leads up to this structure, adding depth to the scene. To the right, a sign reading \"Fountain Service\" is partially visible, hinting at the presence of a water feature nearby. Further back, a snow-covered roof and some trees can be seen, completing the wintry landscape. The overall atmosphere is calm and quiet, with no visible movement or activity, emphasizing the stillness of the"
+    ],
+    "video_ids": [
+        "4e98ca4efe5e096773b086c4cb4fbae934951ea1772e07fde8c69119f8f74e06"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Black plastic container, wooden structure with railing, staircase, sign reading 'Fountain Service', snow-covered roof, trees? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a black plastic container (likely a sled or tub) in the foreground, a wooden structure with railing and staircase in the background, a sign reading 'Fountain Service', snow-covered surfaces, and trees in the distance. All elements described are present and consistent with the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a black plastic container placedlikely a trash) container) placed the foreground, a wooden structure with a and staircase in the background, a sign that 'Fountain Service' and-covered roof, and trees. the distance. The elements match in present and match with the video depicted}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..88e5ce95f40034c985aca7a7c2fda56c1320361f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d1d557d3e2a751969d9791ccc1ae5ea6fd58d8c6eaf0d45da3beabcbb979433
+size 109784
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..10cdae3d42ca3c5f996f196f8784965eac7312ca
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a thrilling ride on a roller coaster. The first frame shows the roller coaster track curving upwards, with the sky visible in the background. The second frame shows the roller coaster track curving downwards, with the ground visible in the background. The third frame shows the roller coaster track curving upwards again, with the sky visible in the background. The roller coaster track is made of metal and is painted black. The roller coaster cars are not visible in the video. The video is shot from a first-person perspective, giving the viewer a sense of being on the roller coaster. The video does not contain any text or other objects. The style of the video is dynamic and exciting, capturing the thrill of the roller coaster ride."
+    ],
+    "video_ids": [
+        "2C7fLcwJ7i4_31_0to158"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Roller coaster track, roller coaster cars (not visible)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a roller coaster track, specifically a section that appears to be an ascent ramp or lift hill, with water dripping from above, which is consistent with a water park or water ride. The track is clearly visible and matches the description of a roller coaster track. Roller coaster cars are not visible, which aligns with the condition as stated.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a roller coaster track with which the section of includes to be part inverted or, loop hill, which amarks from the, which is a with the roller ride ride themed-themed. The track is clearly visible, there the description of a roller coaster track. There coaster cars are not visible in but aligns with the condition ' it.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a718de31f1d5de15df4a740789d1149b769e50d5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c97d738008ef03673ed902ca12730f40d6a1fba6fdd48b60913beac1b4720959
+size 409323
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..120fd8f5361c3f0ab90f0b9f815adc2c2741623e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a delicious breakfast meal being prepared and served. The meal consists of a stack of golden-brown pancakes, topped with a generous drizzle of syrup, and a side of crispy bacon. The pancakes are placed on a black plate, which contrasts nicely with the vibrant colors of the food. The bacon is arranged on the side of the plate, adding a savory element to the sweet pancakes. The plate is placed on a wooden table, which provides a rustic and warm setting for the meal. The video captures the process of preparing the meal, from cooking the pancakes and bacon to plating and serving. The overall style of the video is simple and straightforward, focusing on the food and the preparation process. The colors are bright and inviting, making the meal look even more appetizing. The video is likely intended to inspire viewers to try making the meal themselves or to enjoy a similar breakfast at a restaurant."
+    ],
+    "video_ids": [
+        "kR7FOgmNOu8_30_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A stack of golden-brown pancakes, a drizzle of syrup, and crispy bacon.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a stack of golden-brown pancakes with a drizzle of syrup on them, and crispy bacon being placed beside the stack. These elements match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a stack of golden-brown pancakes with a drizzle of syrup being top. which there bacon is held on the stack. The elements match the description provided, the questionObject(s)' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..571e93608e9378424e358cc6a7108bf2f57c0d04
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4c0779411435d907ce60098b4768445f2bb637bb096543a7feba3d1d3abd806e
+size 118719
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..fb6f74fc7f5f9526509bf0ad4fdf83da73ac74f9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red Mercedes-Benz SUV parked on a brick sidewalk. The car is shiny and well-maintained, with a sleek design and a prominent Mercedes-Benz logo on the grille. The car is parked in front of a building with a sign that reads \"Copper.\" The building has a modern architectural style, with large windows and a flat roof. The brick sidewalk is clean and well-maintained, and there are no other vehicles or pedestrians visible in the video. The overall style of the video is straightforward and unembellished, focusing solely on the car and its immediate surroundings."
+    ],
+    "video_ids": [
+        "igncO_CEn_o_66_0to178"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red Mercedes-Benz SUV parked in front of the building.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red Mercedes-Benz SUV parked in front of a building, matching the core description. The camera zooms in on the front of the vehicle, clearly showing the Mercedes-Benz logo and grille, and the SUV is positioned on a paved area in front of a building with visible signage. Additional cars and background elements do not contradict the main subject.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a red Mercedes-Benz SUV parked in front of a building. which the description description. The car angles in on the vehicle of the vehicle, highlighting showing the Mercedes-Benz logo and the, which the SUV is positioned in a paved area in front of a building with large windows. The elements and a elements are not contradict the main focus of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..efd8fa4ab0c6c5fcf94d3f7502a71acb20c335e9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd8622dc98da68da25b66a3ad0be18fd8302330d8b894b2a333221633caf6298
+size 183206
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..7fc6205434f0cfa565f8fa0d773bea35a9c87c06
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car from the perspective of the driver's seat. The car has a sleek design with a black and red color scheme. The dashboard features a touch screen display that shows a map, indicating that the car is equipped with GPS navigation. The steering wheel is black and has a silver logo in the center, suggesting that the car is a luxury model. The seats are upholstered in red leather, adding to the car's luxurious feel. The car's interior is well-lit, with the sunroof open, allowing natural light to flood in. The car appears to be parked, as there is no movement visible in the image. The overall style of the video is sleek and modern, with a focus on the car's interior design and features."
+    ],
+    "video_ids": [
+        "0Da8u---oxc_19_0to101"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dashboard with a touch screen display, GPS navigation, steering wheel with a silver logo, red leather seats.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a car dashboard with a central touch screen displaying GPS navigation, a steering wheel with a silver Jaguar logo, and red leather seats. All elements described in the condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a car's with a touch touch screen display a navigation. a steering wheel with a silver logo logo, and red leather seats. The the match in the ' are present and match depicted in}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a53c7c60d423e0255d76a45a823308fed7a6c62f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50fdfcffab50383377ed7757563d4510ffbc499a93cf5c3820f5fb100132c063
+size 186630
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..4bbeb47588d641217dbe650eeacc07a1d19a651d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video begins with a close-up shot of a hand holding a clear plastic bag containing a printed image of a cartoon character. The character is depicted in black and white, standing upright with one leg slightly bent and the other extended forward, as if in motion. The character has large, pointed ears and a tail, suggesting it might be an anthropomorphic animal or creature. The hand gently moves the bag, allowing the viewer to see the image more clearly from different angles.  As the scene progresses, the camera zooms out slightly, revealing that the plastic bag is placed inside a larger transparent container filled with water. The container is positioned on a tiled surface, which appears to be part of a crafting or work area. The background includes a green cutting mat with grid lines, indicating a workspace for detailed tasks such as crafting or art projects. A roll of tape and some other small items are visible on the right side of the frame, further emphasizing the crafting environment.  The hand continues to hold the plastic bag"
+    ],
+    "video_ids": [
+        "0eb355617a1884e63c722ab0d3c487320d6dc1f52de814b7adce4080c0a1687c"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A hand, a clear plastic bag, a printed image of a cartoon character, a green cutting mat, grid lines, a roll of tape, and some small items.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a hand holding a clear plastic bag with a printed cartoon character inside. In the background, a green cutting mat with grid lines is visible. Although a roll of tape and some small items are not clearly identifiable or prominently featured, their presence is not contradicted by the visual evidence, and the core described elements are all present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a hand holding a clear plastic bag with a printed image character inside. The the background, there green cutting mat with grid lines is visible, There the roll of tape and some small items are not explicitly visible in in featured, they presence does not necessaryicted by the description content. and they core elements elements are present present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3a464000b95b1f540c4051eb711a57b0813f23b7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1343-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d3f6f168df9eab0e08f702ec08042652e6fd497c39030c40dfcf6d33247fff5
+size 253541
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..a802bdb50d9696302e2d032c8d82c194432af10f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman performing on stage. She is wearing a black outfit with a plunging neckline and a silver necklace. Her blonde hair is styled in an updo, and she has her arms raised in the air, with her fingers pointing upwards. The stage is illuminated with bright lights, creating a dramatic and energetic atmosphere. The woman appears to be singing or speaking into a microphone, which is visible in the image. The style of the video suggests that it is a live performance, possibly a concert or a music video. The focus is on the woman and her performance, with the stage lights and background serving to highlight her presence and the intensity of the moment."
+    ],
+    "video_ids": [
+        "6iDaFQhbDio_3_0to129"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman performing, wearing a black outfit with a plunging neckline and a silver necklace. Her blonde hair is styled in an updo. There is also a visible microphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman performing on stage, wearing a black outfit with a plunging neckline and a prominent silver necklace. Her blonde hair is styled in an updo, and she has a visible microphone attached near her mouth. All elements of the description are accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman performing, stage, wearing a black outfit with a plunging neckline and a silver silver necklace. Her blonde hair is styled in an updo, and she is a microphone microphone in to her mouth. The these of the description are present represented in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..29cc3c43e8cb919bf269acd882b9f09b633b4bdd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:444278517360cf055a23542fcf9f82be7ca6359b95db4934c4e84b6ed63a678d
+size 231504
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..167b3fcbef8f9b94b7bf0090ae7fa1c5b31976c2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a casual, personal vlog shot in a home kitchen. A woman with glasses and a blue shirt is the main subject. She is smiling and looking directly at the camera, giving a friendly and approachable vibe. The kitchen has wooden cabinets and a sink in the background. There are various kitchen items and decorations, such as a potted plant, a vase, and a bottle. The lighting is soft and warm, suggesting an indoor setting. The style of the video is informal and personal, likely intended for a social media platform or a personal blog."
+    ],
+    "video_ids": [
+        "13omj9863Ko_44_0to173"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with glasses and a blue shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman wearing glasses and a blue shirt, which matches the core description. Additional elements like the kitchen background do not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a woman wearing glasses and a blue shirt. which matches the description description. The elements such the kitchen background and not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7373c632d1b41918a890d0869c017d4d85ecff77
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a3ba2baebdf6a1449f2cd8d0328009ba645088ad2035796b89850809ad29b23
+size 237325
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..c84356ebc3daa3c73c76f1f287e50dbc49fc39fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a cooking tutorial, featuring a wooden cutting board with various ingredients and a glass bowl. The ingredients include sweet potatoes, green peppers, and cherry tomatoes. The style of the video is simple and straightforward, focusing on the preparation of a healthy and nutritious meal. The cutting board is the main subject of the video, with the ingredients arranged neatly around it. The glass bowl is used to hold the sweet potatoes, and the green peppers and cherry tomatoes are placed on the cutting board. The video is likely to be informative and educational, providing viewers with a step-by-step guide to preparing a delicious and healthy meal."
+    ],
+    "video_ids": [
+        "hHTCykx101w_5_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Wooden cutting board, sweet potatoes, green peppers, cherry tomatoes, glass bowl? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a wooden cutting board with sweet potatoes, green peppers, cherry tomatoes, and a glass bowl as central elements. These items are prominently displayed and match the described objects without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows shows a wooden cutting board with a potatoes, green peppers, cherry tomatoes, and a glass bowl. the elements. The objects match present displayed, match the description objects. any or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..05b53300b500dbeebcfd906a076efbb796dc826a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:94ea2b3f415679d93c13c54807c5c059a0c42ef54dd3855a43be456738b8858a
+size 80006
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..27f18f0fce4bad93ec827461517a14702abc3ffc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen preparing a meal on a wooden table. The person is placing green leaves into a glass jar with a metal clasp. The jar is filled with water and has a few pieces of fruit inside. The person is wearing a ring on their finger. In the background, there is a bowl of noodles. The style of the video is a close-up shot of the person's hand and the jar, focusing on the action of adding the green leaves to the jar. The video captures a moment of food preparation, emphasizing the use of fresh ingredients."
+    ],
+    "video_ids": [
+        "_9rZkhrhINM_5_0to160"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person, glass jar, green leaves, fruit, ring, bowl of noodles? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a person's hand placing green leaves into a glass jar, which already contains sliced mushrooms and liquid. The hand has a ring, and a bowl of noodles is visible in the background. All specified objects are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a person placing hand placing green leaves into a glass jar, which contains contains some fruit. a. The presence is a ring on and in bowl of noodles is visible in the background. The the objects are present and correctly depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6a9691a480a95872dd3837b07fddb86fa63f3b71
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7fc100bc2dfcf95f57291f47e8f168148dc90bbbb6094e304fc622292cdf5e67
+size 141643
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..5543529000f712c9dd045fa8716d7c082ac6250d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a small monkey with a surprised expression, captured in three frames. The monkey has a furry body, large eyes, and a small nose. In the first frame, the monkey is looking to the left with its mouth slightly open. In the second frame, the monkey's mouth is open wider, and it appears to be making a sound. In the third frame, the monkey's mouth is closed, and it is looking directly at the camera. The background consists of a blurred image of a wicker chair and a window, suggesting an indoor setting. The style of the video is a close-up, candid shot of the monkey, capturing its expressive face and reactions."
+    ],
+    "video_ids": [
+        "NTa_4g8eVDk_79_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small monkey with a surprised expression.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small monkey that initially has a surprised expression, with wide eyes and an open mouth. Although the monkey later turns its head and moves, the core description of a small monkey with a surprised expression is largely fulfilled in the beginning of the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small monkey with appears appears a surprised expression, which wide eyes and an open mouth. The the video's appears its head and the its the initial description of a small monkey with a surprised expression is largely fulfilled. the first of the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6e0acc980b553c1ad6518975d4bbd16ae1cec393
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:25bd3b375c4a64a0ca2f7918ae550c40522ff1839155cac1abb3ac2051fc324d
+size 173208
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..f4c7f97939e9a7347194188b04255be953be31ec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a blue pickup truck parked in front of a building with a red brick facade. The truck is a four-door model with a black grille and black wheels. The vehicle is positioned at an angle to the camera, allowing a clear view of its side profile. The building behind the truck has a large garage door and a smaller door on the right side. The sky is clear and blue, suggesting a sunny day. The style of the video is a straightforward, real-life depiction with no special effects or artistic filters. The focus is on the truck and its immediate surroundings, with no additional action or movement."
+    ],
+    "video_ids": [
+        "MNdKf3NT6P8_28_0to153"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue pickup truck and a building? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a blue pickup truck parked in front of a building, which matches the 'Object(s)' condition. The truck is prominently featured, and the building serves as the background setting. No elements contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a blue pickup truck parked in front of a building with which matches the descriptionObject(s)' condition. The truck is prominently featured in and the building, as the backdrop,. There additional contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0be4a26c0336b1b6b06605424d720a00508be309
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c20e4e1b665402695a6ccbff9f76d54a2603f0fd8e740bcd381f3a43769bc4a
+size 202822
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..5cfd9aff103a8a6ce2733099d0dacfb5c8a3737e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is an aerial view of a residential area with a large white building in the center. The building has a flat roof and is surrounded by trees. The surrounding area includes other buildings with different roof styles and colors. The road in front of the building is paved and leads to a body of water in the background. The sky is clear and blue, suggesting a sunny day. The style of the video is realistic and it captures the details of the buildings and the landscape."
+    ],
+    "video_ids": [
+        "rHbcxH3-XPg_112_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Large white building, trees, other buildings, paved road, body of water? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a large white building, surrounding trees, other buildings, a paved road, and a body of water. All these elements are visible and consistent with the description, even though the video is AI-generated.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a large white building, numerous trees, other buildings, a paved road, and a body of water in The these elements are present and match with the description provided indicating though the video is a-generated,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b6061b1b5886114f262bdeef77190bb72a51082d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:379dce0952895492676bb89a08e63ee4c3aa356b920114526bcb2431790c7aad
+size 185105
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..2d258b6e992e6db25747a21ab014d68240d1b4e3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a cat lying down on a wooden surface. The cat appears to be relaxed and comfortable, with its eyes closed and whiskers spread out. The cat's fur is a light orange color, and it has a blue collar with a tag. The cat's position changes slightly between frames, but it remains in a relaxed state. The wooden surface provides a warm and natural backdrop for the cat. The style of the video is a simple, unobtrusive close-up that focuses on the cat's peaceful demeanor."
+    ],
+    "video_ids": [
+        "GTVIuS2RIFk_4_0to102"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A cat (relaxed, with closed eyes and spread-out whiskers), a wooden surface, a light orange fur, a blue collar with a tag.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a cat lying on a wooden surface with its eyes closed and whiskers spread out, matching the description. The cat has light orange fur and is wearing a blue collar with a tag, which is clearly visible. All elements described are present and consistent with the video content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a cat with on a wooden surface with its eyes closed and whiskers spread out. which the description. The cat has light orange fur and is wearing a blue collar with a tag, which also also visible. The the of in present and accurately with the video content.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..06002f55f18e085e04c257ba871f9ca93d912a66
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1344-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:42db01a75e7ebfab110fdf335d7b2e4e38491b28e4f5d08e4ec10a08b502b78e
+size 282790
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..9389922c50dab9c483ef8b1b34a2b700a0081640
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man standing next to a black pickup truck in a parking lot. The man is wearing a red shirt and a beige vest, and he is gesturing towards the truck with his right hand. The truck is an older model, with a chrome grille and a black bed liner. The parking lot is empty, and the sky is clear and blue. The man appears to be explaining something about the truck, possibly its features or history. The style of the video is straightforward and informative, with a focus on the man and the truck."
+    ],
+    "video_ids": [
+        "JrviCx1kS8Q_21_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a black pickup truck? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man standing next to a black pickup truck, which matches the 'Object(s)' condition. The man is visible on the right side of the frame, and the black pickup truck occupies the left and central portions of the frame. There are no elements that contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man standing next to a black pickup truck. which ful the 'Object(s)' condition. The man is interacting and the right side of the frame, and the black pickup truck is the left side central part of the frame. The are no additional in contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..573a980110ca1a39f4745fe17f46d41395a04957
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac6f52157d295e9c720a29aac122c6ec2dff618e28200ad0d9b6c1a90816b8c8
+size 170361
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..460d20084836124f98050c13f8c95b44091e50fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a close-up of a tiger's face, showcasing its detailed fur pattern and expressive eyes. The tiger appears to be in a natural setting, with a blurred background that suggests a forest or jungle environment. The lighting in the video is soft and natural, highlighting the tiger's features without creating harsh shadows. The style of the video is realistic, with attention to detail in the tiger's fur and facial features. The camera angle is slightly above the tiger's head, giving a sense of the animal's size and power. The video does not contain any text or additional elements, focusing solely on the tiger's face and expression."
+    ],
+    "video_ids": [
+        "lCbOUwCQc-8_13_0to117"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Tiger's face? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video focuses on a close-up of a tiger's face, showing detailed expressions and features such as eyes, nose, whiskers, and mouth. The tiger's face is the central and dominant element throughout the frames, fulfilling the 'Object(s): Tiger's face' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows on a close-up of a tiger's face, capturing detailed features and fur such as the, nose, anders, and fur. The image's face is the central and primary element in the frames, fulfilling the 'Object(s)' Tiger's face' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..58a0f6a94b4b7c52ce082aa0c12bdb455fd13f24
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eaa0344f1183fef1e06d65bbf2a84bd3f82eee3e09e366d931b7935ebf4acfab
+size 290747
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..3cccb3f6b97b0f84b371a1e511cd87304fe22d91
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a small, rectangular aquarium placed on a wooden surface against a dark background. Inside the aquarium, there is a vibrant scene featuring colorful pebbles at the bottom, a red decorative rock, and green aquatic plants. A goldfish swims gracefully among these elements, adding life to the setup. To the right of the aquarium, a clear plastic bottle is positioned upside down on a gray base. The bottle is connected to the aquarium via a transparent tube, which allows water to flow from the bottle into the tank. Bubbles rise from the bottom of the bottle, indicating the ongoing process of water transfer. The video captures the serene interaction between the fish and its environment, emphasizing the simplicity and functionality of the setup. The \"Craft Master\" watermark appears in the bottom left corner throughout the video, suggesting it might be part of a series or tutorial."
+    ],
+    "video_ids": [
+        "b96a3db89245544ce33452c38138a1921ecb931a98e20e662826bc64b9afe1aa"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Aquarium, colorful pebbles, red decorative rock, green aquatic plants, goldfish, clear plastic bottle, gray base, transparent tube, bubbles.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows an aquarium with colorful pebbles at the bottom, a red decorative rock, green aquatic plants, and goldfish swimming inside. To the right, there is a clear plastic bottle with a gray base, and a transparent tube connected to the bottle and aquarium, with visible bubbles rising in the water. All listed elements are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts an aquarium with colorful pebbles, the bottom, a red decorative rock, and aquatic plants, and afish swimming inside. There the right, there is a clear plastic bottle with a gray base and a a transparent tube is to it bottle, the, which bubbles bubbles indicating from the tube. The elements elements are present and match with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ca87269140b06ad463fd55868ae0d6bfbb4f2d10
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a2615f91329617cca878a53a69bd17c7f1d773919dfe14f6b3dba7c3d05dac2b
+size 104624
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..0a30b6768a8def464eb91d48bd16ccbe9287c3c9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a casual conversation between two men, one standing and the other sitting on a motorcycle. The standing man is wearing a red shirt and is smiling, while the man on the motorcycle is wearing a black and white jacket with a red logo. The motorcycle is parked on a road with a grassy field in the background. The sky is overcast, suggesting it might be a cool or cloudy day. The style of the video is informal and friendly, with the two men engaging in a relaxed conversation."
+    ],
+    "video_ids": [
+        "FUopO25QQ8w_47_27to205"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, one standing and one sitting on a motorcycle. The standing man is wearing a red shirt and smiling. The man on the motorcycle is wearing a black and white jacket with a red logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men: one standing in a red shirt, smiling, and the other sitting on a motorcycle wearing a black and white jacket with a red logo. These elements match the description exactly, with no contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men, one standing and a red shirt and smiling, and the other sitting on a motorcycle wearing a black and white jacket with a red logo. The elements match the description provided, with no additional or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..22f9c2c00959352b21b47bf897b04c8957f55977
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bad1bfc1414bc98ebe14d5eafcfcd3f4fb1ddbd4c5310dabf0e23516a432ca71
+size 129939
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..291c4abad68cfda546339a9643eca9eb19a62733
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a scene at a car show, featuring a variety of vehicles on display. The main focus is a gold-colored muscle car, which is parked prominently in the foreground. The car's shiny exterior reflects the bright sunlight, highlighting its sleek design and powerful presence.  In the background, there are several other cars, including a silver sports car and a classic white station wagon. These vehicles are also on display, adding to the diversity of the car show.  The setting is a large parking lot, which is filled with cars and people. The lot is expansive, with plenty of space for the vehicles and attendees. The sky above is clear and blue, suggesting a beautiful day for the car show.  Overall, the video provides a glimpse into the world of car enthusiasts, showcasing the beauty and power of these machines in a vibrant and lively setting."
+    ],
+    "video_ids": [
+        "HiAqR2aK210_17_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A gold-colored muscle car, a silver sports car, and a classic white station wagon.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a gold-colored car in the foreground, which appears to be a muscle car (likely a Buick Regal or similar model from the 1980s). Behind it, there is a silver car with its hood open, which looks like a classic sports car (possibly a Chevrolet Camaro or similar). Also visible is a classic white station wagon in the background. All three vehicles match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a gold-colored muscle that the foreground, which appears to be a muscle car,likely a Dodgeick Skylal, a model). the 1970s). Behind it, there is a silver sports that a hood up, which could like a sports sports car.possibly a Chevrolet Camaro). similar). Additionally visible is a classic white station wagon in the background, The these vehicles match the description provided,}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c98a444d9b85727f61b4089bee08a8d72b8c0a89
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8625cdc6a6a9a4531c03baeed18b1c44fe97ef68675995f889a85b385032c9b9
+size 225894
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..a7d2311edb30f36eb804b0232d83f638a564ee47
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two white racing cars parked in a garage with a graffiti-covered wall in the background. The cars are positioned side by side, with the one on the right being larger and more prominently featured. The garage has a high ceiling and a concrete floor, and the lighting is bright, illuminating the cars and the wall behind them. The graffiti on the wall is colorful and includes various designs and text. The style of the video is realistic, with a focus on the cars and the garage setting. The video likely captures the cars in a moment of rest or preparation before a race, as they are parked and not in motion. The graffiti on the wall adds a touch of urban art to the otherwise industrial setting."
+    ],
+    "video_ids": [
+        "WeIkNhBSiHU_6_0to117"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two white racing cars, a graffiti-covered wall? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two white racing cars prominently positioned on a platform, and behind them is a wall with visible graffiti, including a skull and colorful spray-painted designs. These elements match the 'Object(s)' condition described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two white racing cars parked displayed in a concrete, with behind them is a wall covered vibrant graffiti. which colorful large and other abstract paintainted designs. The elements match the descriptionObject(s)' condition described in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cc1c6992d19b0ca473c93478364c918323b325c6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d8408036a19d483019947d52138fe345bff3a87f73d5b1f0ebff4785dfd73fd0
+size 208422
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..bf58306eac0efbcdb15aea94998eb73f0a417514
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a car driving on a snowy road. The car is a dark-colored SUV with a visible spare tire on the back. The road is covered in snow, and the car's tires leave tracks in the snow as it moves. The surrounding environment is a mix of snow-covered ground and trees, suggesting a winter setting. The car's headlights are on, illuminating the snowy road ahead. The style of the video is realistic, with a focus on the car and its movement through the snowy landscape. The camera angle is from the side of the car, providing a clear view of the vehicle and its surroundings."
+    ],
+    "video_ids": [
+        "d9hRz3q7hIU_63_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A dark-colored SUV with a spare tire, snow-covered ground, and trees.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dark-colored SUV driving on snow-covered ground, with trees visible in the background. The SUV's tires are clearly visible, and the snow-covered terrain and surrounding trees match the description. Although no spare tire is visibly attached to the vehicle, the core elements described are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a dark-colored SUV with on a-covered ground. which trees visible in the background. The vehicle has spare are partially visible, and one presence on ground is trees trees match the description. The the spare tire is directly mounted to the vehicle, the presence elements of in present,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..331825064ff997aa152af529a292550fe181fb36
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:541409c1bcd492784b6f623725d831097929d57c334a3e7d10dd753c0253d088
+size 277274
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..bad948bbcd6b3a3eff9483c9c45a283f982279c3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a baseball game in progress. The scene is set on a baseball field with a well-maintained diamond and lush green grass. The players are dressed in traditional baseball uniforms, with the home team wearing white and the visiting team in blue.  In the first frame, the batter is in the midst of a powerful swing, his body coiled with anticipation. The pitcher, standing on the mound, watches the trajectory of the ball with a focused gaze. The umpire, positioned behind the batter, is ready to make a call.  In the second frame, the ball is in mid-air, having just been hit by the batter. The fielders are alert, their bodies poised to react to the ball's trajectory. The batter, having completed his swing, is starting to run towards first base.  In the third frame, the ball has landed in the outfield, and the fielders are in motion, running towards it. The batter, now a base runner, is sprinting towards first base, his eyes on the prize. The pitcher, having thrown the ball, is watching the play unfold.  The video is a dynamic snapshot of a baseball game, capturing the tension and excitement of the sport. The players' movements and the trajectory of the ball create a sense of action and anticipation. The video is shot from a perspective that places the viewer right in the middle of the action, making them feel like they are part of the"
+    ],
+    "video_ids": [
+        "gctJp_I258I_48_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Batter, pitcher, umpire, fielders? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a baseball game scene with a batter at the plate, a pitcher (partially visible in the foreground), an umpire behind the catcher, and fielders in the background. All core elements mentioned are present, even though the pitcher is partially obscured by another player in the foreground.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful the baseball game with with a batter, the plate, a pitcher throwingthoughially visible), the background), an umpire, the batter, and fielders in the background. The the elements of in present, fulfilling though the pitcher is not obscured. the player. the foreground.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5e67da0c072fdfde0ad41f1a7920bf705fcf6128
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1345-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88e014d23cda0cf9f99ab3f4c9f7157107f51da2d9c82955815ea5a9cbf5d2d1
+size 291218
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..8acb593b08e82fe61c0f642a757dc916eeb637d1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red BMW X4 parked in a snowy environment. The car is covered in a layer of snow, with the rear lights glowing red. The license plate is visible and reads \"CENW00.\" The car is parked on a snow-covered road, and there is a fence visible in the background. The style of the video is realistic, with a focus on the car and its surroundings. The lighting is natural, and the colors are muted due to the snow. The video captures the car in a still position, with no movement or action taking place."
+    ],
+    "video_ids": [
+        "8kuxfLfJYUo_9_0to112"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red BMW X4? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red BMW X4, identifiable by its distinctive rear design, the 'X4' badge, and the BMW logo. The car is covered in snow, which is consistent with a winter setting, and no elements contradict the core description of the vehicle.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a red BMW X4, identifiable by its distinctive rear design, including BMWBMW4' badge, and the overall logo. The car is parked in snow, indicating is a with the winter setting, and the other contradict the description description of the object.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2bb74deec8c1d516338fdb73c22fd13c6cea5c49
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59163c34f4cacb030e44bc9810264903812121b06e1d12bf39f58487c0c48a7e
+size 153847
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..e271d9b7c6d262301c378073cf248db95fe54ce9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man on a boat, wearing a straw hat and an orange shirt. He is steering the boat, which is on a body of water. The man appears to be enjoying his time on the water, and the boat is moving smoothly. The water is calm, and the weather seems to be sunny. The man is the main subject of the video, and there are no other people or objects in the frame. The style of the video is casual and relaxed, capturing a moment of leisure and enjoyment."
+    ],
+    "video_ids": [
+        "m3HLWpSdUFE_35_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man on a boat, wearing a straw hat and an orange shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man on a boat, wearing a straw hat and an orange shirt, which matches the description. The presence of the boat's controls and the surrounding water does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person on a boat, wearing a straw hat and an orange shirt, which matches the description provided The background of the boat, interior and the water water and not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3eb977a268ff20cbcf46373a0e02f23185a99d06
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb6f277e796d9151ec0baf24fd5ba784f5f5fa63ca5cbcfa93de2da0ba0892a2
+size 201630
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..11b5f640f71182bd64ec543a7a52f35a797d427b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment on a football field. A player, dressed in a white and orange uniform with the number 22, is the main focus. He is in the midst of a celebratory gesture, flashing a peace sign with one hand and sticking out his tongue with the other. His energy is palpable, reflecting the excitement of the game. The background is filled with a crowd of spectators, their faces a blur of anticipation and excitement. The atmosphere is electric, the crowd's energy feeding into the player's performance. The video is a snapshot of a thrilling moment in a football game, capturing the player's triumph and the crowd's reaction."
+    ],
+    "video_ids": [
+        "I0__0eHxafM_13_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a white and orange uniform with the number 22.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a white and orange uniform with the number 22 and the word 'CLEVELAND' on the jersey, which matches the description. The player's pose and the background elements (crowd, stadium) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a player player wearing a white and orange uniform with the number 22. the number '2VELAND' on the jersey. which align the description. The player is attire and the background crowd,crowd and field) are not contradict the core description of}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e6a53b156a6415701e7456f7c323d2464d236310
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7cd3f55529ea030f1a2e89c6348fc1a162d72fd9a4876250bea918d871bca81e
+size 234182
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..e964cb955d20f8ef565dc0803c2b10739754abe5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen playing a black acoustic guitar. The individual is wearing a blue jacket and a baseball cap. The guitar is held in a playing position, with the person's hands on the neck and strings. The setting appears to be a living room, with a couch and a coffee table visible in the background. The room has a modern decor, with a painting hanging on the wall. The overall style of the video is casual and relaxed, capturing a moment of leisure and enjoyment."
+    ],
+    "video_ids": [
+        "QqNdkNZHffE_12_28to206"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person playing a black acoustic guitar, wearing a blue jacket and a baseball cap.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person holding and playing a black acoustic guitar, wearing a blue jacket and a baseball cap. The core description is accurately represented, even though there are additional elements like a mustache, glasses, and a decorative wall sculpture, which do not contradict the main condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a person playing and playing a black acoustic guitar. which a blue jacket and a baseball cap. The guitar elements is largely represented in with though the is no elements like the paintingache and which, and a different background in in which do not contradict the main condition.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1ddea78c986099a301362227c4761dcf0515e9f9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1473fd19a06f63105869a99791a9df01189190b828f2623bddb1a0718bb7072
+size 240059
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..c66273c498852746c3a6f595aa1887fb6a556af1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a casual dining experience between two men at an outdoor restaurant. The men are seated at a table, each with a plate of food in front of them. They are engaged in conversation, enjoying their meal and each other's company. The restaurant is set in a tropical location, with lush greenery and palm trees visible in the background. The atmosphere is relaxed and inviting, with the sound of birds and the rustling of leaves adding to the ambiance. The men are dressed casually, suggesting a laid-back and informal setting. The video is shot in a documentary style, capturing the essence of the moment without any artificial staging or embellishment."
+    ],
+    "video_ids": [
+        "7DSBk77gHKQ_47_0to147"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men seated at a table with plates of food in front of them.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men seated at a table with plates of food in front of them, which matches the core description. Additional elements like background activity or minor visual artifacts do not contradict this primary depiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two men seated at a table with plates of food in front of them. which align the description description. The elements such the plants and other details details do not contradict the primary condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e9d5b74f78a902fbffef08f59371336097b65df0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b7791c440ebfa8f97acfd26bc483ae46c1c3b55967e4f02b2662fbcc7dfda69
+size 163341
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..2170452c568319068da7451181296d2cc8279897
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features Dr. Ralf Speth, CEO of Jaguar Land Rover, standing in front of a white sports car. The car is parked in a garage with neon lights and a blue and white sign that reads \"Jaguar\". Dr. Speth is dressed in a brown suit and a white shirt, and he is gesturing with his hands as he speaks. The video appears to be a promotional or informational video about Jaguar Land Rover, with Dr. Speth likely discussing the company's products or plans. The setting suggests a modern and stylish environment, fitting for a luxury car brand."
+    ],
+    "video_ids": [
+        "KHth-iQdN2Q_10_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dr. Ralf Speth, a white sports car, neon lights, a blue and white 'Jaguar' sign.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing Dr. Ralf Speth, a white sports car (specifically a Jaguar I-PACE), neon lights in the background, and a blue and white 'Jaguar' sign. All specified elements are present and accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as featuring showing a. Ralf Speth, a white sports car,aally a Jaguar),-PACE), neon lights, the background, and a blue and white 'Jaguar' sign. The these elements are present and match depicted in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5deac5d4edd96b8556bfb284546d6970ded05bf0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:553f4c9f2e887779141ae5a89e9ad4c30971356e8334714382308f44532be673
+size 109774
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..3b4caa0c7eeca74f60112f62d789e65b0ad35f95
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a motorcycle wheel in three different stages of motion. The style of the video is a time-lapse or slow-motion sequence, capturing the intricate details of the wheel's movement. The motorcycle wheel is black with a silver hub and spokes, and the tire has a pattern of grooves and ridges. The background is blurred, emphasizing the wheel as the main subject. The video is likely intended to showcase the motorcycle's design and engineering, highlighting the precision and craftsmanship of the wheel's components."
+    ],
+    "video_ids": [
+        "1yLFa-tRheE_17_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Motorcycle wheel? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a close-up of a motorcycle wheel, including the tire, spokes, disc brake, and chain sprocket. These elements clearly identify the object as a motorcycle wheel, fulfilling the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a close-up of a motorcycle wheel, showcasing the tire, rim, and brake, and part guardprocket. The elements are indicate the object as a motorcycle wheel, fulfilling the ' specified}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fcc8fba4e4624f7138ce74d1bf6d5d035727d08c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5591219f325289fb9c07b5f008467819681c6e1edd3d6159a76a8d1146416ad1
+size 389534
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..6d7f577c339f2b120251161352844ac14fbcf0ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in front of a large screen displaying two news anchors. The man is dressed in a black leather jacket and a striped shirt, and he is pointing towards the screen with a smile on his face. The screen is purple and displays the CNN logo, indicating that the news anchors are from CNN. The man appears to be presenting or discussing the news anchors, possibly in a humorous or satirical manner. The overall style of the video suggests a light-hearted or comedic tone."
+    ],
+    "video_ids": [
+        "JHOgUkuSXn4_19_25to191"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a large screen displaying two news anchors.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man standing in front of a large screen that displays two news anchors. The man is visible and the screen is prominently featured, fulfilling the core description. Additional elements like the Netflix branding and studio lighting do not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man standing in front of a large screen that displays two news anchors. The man is positioned and positioned screen behind prominently featured, fulfilling the ' description of The elements such the man logo on the setting do not contradict the core.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e118a273be6e51839d601f9ace4aedf0804ced89
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1346-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30d477ecbf3ba9ec92a1a30743900299eef7a63fac7b88c13a82f7a3f04748ce
+size 204665
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..cc20036c127c69c49baca110ad4dd2ef22404428
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a breathtaking view of a city skyline during sunset. The sky is painted with hues of orange and pink, creating a warm and inviting atmosphere. The cityscape is dotted with tall buildings, their silhouettes standing out against the vibrant sky. The Space Needle, a prominent landmark, is visible in the distance, adding a touch of modernity to the scene. The sun is setting, casting a golden glow over the entire scene. The video is taken from a high vantage point, providing a panoramic view of the city and its surroundings. The overall style of the video is serene and picturesque, capturing the beauty of the city at sunset."
+    ],
+    "video_ids": [
+        "E6qnfmJZ0_w_17_0to178"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Tall buildings, the Space Needle? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features the Space Needle, a tall iconic structure, along with other tall buildings in the urban skyline. These elements are clearly visible and match the 'Object(s)' condition described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts features tall Space Needle, which well and structure, which with other tall buildings in the city skyline. The elements are clearly visible against align the descriptionT(s)' condition described.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..361d83374f9884e47b16c717f3e1af9157347db9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8beb50460dca8b3ba6b2122821f1bcae6f9137a1fe7ef9a64f00f46fbef4f1db
+size 84166
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..593938db488f3afa0fbb513718f9aaa949e6e480
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen preparing a smoothie. The first frame shows the person holding a ripe banana over a glass filled with a brown liquid, which appears to be a smoothie. In the second frame, the person has begun to peel the banana, with the peel partially removed. In the third frame, the person is in the process of adding the banana to the smoothie, with the peel now fully removed. The smoothie is in a clear glass, and the person is wearing a yellow shirt. The style of the video is a simple, everyday cooking or food preparation scene, with a focus on the action of adding the banana to the smoothie."
+    ],
+    "video_ids": [
+        "MIjDB8dOkto_9_0to159"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person, ripe banana, glass filled with smoothie? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person holding a ripe banana and dipping it into a glass filled with a brown smoothie. The core elements described \u2014 person, ripe banana, and glass filled with smoothie \u2014 are all present and clearly visible, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person interacting a ripe banana above placing it into a glass filled with a smooth smoothie. The presence elements of\u2014 a, ripe banana, and glass filled with smoothie \u2014 are all present and accurately depicted in fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..284c601407e7d0b390c057447cef9176653d0458
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f72a0723c44c0cb3bbfade256601a63ab10164f1747711d421e17964ee03988a
+size 155544
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..84cf184305661af6a8611fde2919c48ad3e069b5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a floral shirt and orange glasses, standing in a kitchen. He is looking to the side with a slight smile on his face. The kitchen is equipped with a refrigerator and a clock on the wall. The man's attire and the kitchen setting suggest a casual and relaxed atmosphere. The video captures a moment of everyday life, with the man appearing to be in a good mood. The kitchen setting implies that he might be preparing to cook or just finished cooking. The clock on the wall indicates the passage of time, adding a sense of realism to the scene. The overall style of the video is simple and straightforward, focusing on the man and his surroundings without any additional embellishments."
+    ],
+    "video_ids": [
+        "ksDeXH403cM_25_22to240"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a floral shirt and orange glasses, standing and looking to the side with a slight smile.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a floral shirt and orange glasses, standing and looking to the side with a slight smile. These elements are consistent with the description provided, even though he also gestures and turns his head slightly throughout the clip, which does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a colorful shirt and orange glasses, standing and looking to the side with a slight smile. The elements match consistent with the description provided. making though the is appears towards moves his head slightly, the video, which is not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..13ae7fcbe7926766cfea1e3e941d22448ff081f8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8eaf85ad7dbe93e7033cbd444b98722efe3cd1d491e4b97782825e6f99ac2e63
+size 290422
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..841311a8f9faf77970c2a4e3240a1680578b63f5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a male news anchor. The anchor is dressed in a suit and tie, standing in front of a large screen displaying the news channel's logo and a weather forecast. The screen also shows a graphic with the text \"Tomorrow will be\" followed by a blank space, suggesting an upcoming weather forecast. The newsroom has a professional and modern look, with multiple monitors and a desk visible in the background. The style of the video is typical of a news broadcast, with the anchor delivering news and information to the viewers."
+    ],
+    "video_ids": [
+        "h8zgPcib6aE_11_16to205"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A male news anchor, a large screen, multiple monitors, and a desk.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a male news anchor standing in front of a large screen displaying weather information. Multiple monitors are visible in the background, and a desk can be seen to the right side of the frame with a person seated at it. All core elements mentioned in the condition are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a male news anchor standing behind front of a large screen. a information. There monitors are visible on the background, and the desk is be seen in the left of of the anchor. a yellow seated at it. The elements elements of in the condition are present and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8e278bdcc22463bef29a1aeb158005df3be42ed0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:39620de75a243b43b312ec4974e51ec0db09d8b9018a8d78d5214a7a778eb7ba
+size 94770
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..8082675dc1152e08ed25fe1cc1567b8fad43e3be
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a river flowing through a lush green valley. The river, with its calm waters, is surrounded by verdant grass and trees, creating a tranquil atmosphere. In the distance, a small village with houses and a church can be seen, adding a touch of human presence to the otherwise natural landscape. The sky above is a clear blue, dotted with fluffy white clouds, completing the picturesque setting. The video is shot from a high angle, providing a panoramic view of the valley and the river, allowing the viewer to appreciate the beauty of the landscape in its entirety."
+    ],
+    "video_ids": [
+        "NUiJzk_GhEs_1_0to188"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: River, grass, trees, a small village with houses and a church, the sky? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a river flowing through a valley with grass and trees on its banks. On the far side, there are buildings that resemble a small village, including structures that could be interpreted as a church. The sky is visible with clouds, matching the description. There are no significant contradictions with the core elements mentioned.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a river flowing through a landscape, lushy trees on both banks. There one left side of there is houses that resemble a small village, including what that look be interpreted as houses church. The sky is visible and a, fulfilling the description. The are no elements contradictions or the given elements mentioned.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c3fcf5d2cf935cf97b06b659e0f53639baa11584
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63f895ca64fd99ac1a1d098ab4751e5fd019d5fe9f5e35e9a65c83ec404378b1
+size 390592
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..8e5660e097651b0c9e1f4a3579d2597ae0252854
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a car driving down a long, straight road in a desert-like environment. The car is a dark-colored SUV with a roof rack, and it is moving towards the right side of the frame. The road is marked with a double yellow line and a white line on the right side. The surrounding landscape is arid and barren, with sparse vegetation and a clear sky. The style of the video is realistic, with a focus on the car and the road, and it captures the sense of isolation and vastness of the desert environment."
+    ],
+    "video_ids": [
+        "SQtnVdjZlBc_191_0to139"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A dark-colored SUV with a roof rack.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a dark-colored SUV with a visible roof rack driving on a road. The vehicle's appearance and the presence of the roof rack match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a dark-colored SUV with a roof roof rack, on a road. The vehicle matches color and the presence of the roof rack match the description provided. the questionObject(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0d4445285877beddf0cc35d47de8b3b68b482ce3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64e15413f455f48ae800a0623d971447ce3a32152116f0ec0d43953462a7db0c
+size 147939
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..bdfd2b767fe0228a701e60abbd9221c5b283f43d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a robot dressed in a white space suit with a gold helmet, standing in a space-themed environment. The robot is adorned with NASA and GM logos on its chest, indicating a partnership between the two organizations. The robot is positioned in front of a blue wall, which adds a sense of depth to the scene. The overall style of the video is futuristic and sleek, with a focus on the robot as the main subject. The lighting is bright and even, highlighting the details of the robot's suit and helmet. The video does not contain any text or additional objects, and the robot remains stationary throughout the frames. The video is likely intended to showcase the capabilities of the robot and the collaboration between NASA and GM."
+    ],
+    "video_ids": [
+        "llix751qov0_15_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A robot dressed in a white space suit with a gold helmet, adorned with NASA and GM logos.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a robot wearing a white space suit with a gold helmet, and both the NASA and GM logos are clearly visible on its chest. The description matches the visual content without contradiction, even though the background includes additional elements like machinery and equipment, which do not conflict with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a robot dressed a white space suit with a gold helmet, which it NASA NASA and GM logos are visible visible on the chest. The robot align the core elements of any.\"\n and though the robot is a elements that a and a, which are not conflict with the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0b6e25ae7aa47201896f98b3f055d1055b1df1ce
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d8a01e42790a96fb14e714ecd23ea436ad9677584f8a354f73900343a7a83a5b
+size 171178
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..4aad93612c7944276df7cb5f23ed32f7366a2086
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a kitchen, holding a bowl filled with white foamy substance. She is wearing a blue shirt and glasses, and her facial expression suggests surprise or confusion. The kitchen is well-equipped, with a refrigerator, oven, and sink visible in the background. There are also various items on the counter, including a bottle, a vase, and a potted plant. The woman's hand gestures and the context of the scene suggest that she is either preparing a recipe or reacting to an unexpected outcome. The overall style of the video is casual and candid, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "XoWgMy6IOsE_7_61to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a bowl filled with white foamy substance, a blue shirt, glasses, a bottle, a vase, and a potted plant.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing glasses and a blue shirt, holding a bowl filled with a white foamy substance. In the background, a bottle, a vase, and a potted plant are visible, matching the described elements. All core objects are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a and a blue shirt, holding a bowl filled with a white foamy substance. There the background, there bottle, a vase, and a potted plant are visible, which the description elements. The the objects and present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6319d225c79c7a40763cb835012cc83f8479b4ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1347-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a90fc559004546ca05e4709674d26d93b057d879293fafde1800a90ceac8d950
+size 132293
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff717b9a3b4f4a4ee516b9924a16acb3c62620eb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen in a kitchen setting, preparing food. He is wearing a black shirt with a polka dot pattern and has a beard. The man is using a white food processor to blend ingredients. The kitchen is equipped with various appliances, including a refrigerator and a microwave. The man is also wearing a pair of glasses. The video captures the man's actions in a realistic and detailed manner, providing a glimpse into his cooking process."
+    ],
+    "video_ids": [
+        "9FuHvacZbCw_41_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a white food processor, a black shirt with polka dots, a beard, and glasses.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and glasses wearing a black shirt with polka dots. He is interacting with a white food processor on a table. All elements mentioned in the condition are present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a beard and glasses wearing a black shirt with polka dots. He is using with a white food processor in a kitchen. The the in in the condition are present and accurately depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ca18834b37352d0ac0b3bd37a1b242f7a0d361b8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:455b47c05813dbf649a1d9a3388feece806a8dd8067756d402e3369af3de34d1
+size 189498
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..0d6ec4db594d52e75819c4e40b00183ec60d7b6c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen cooking in a kitchen. She is wearing a black top and is standing in front of a stove. She is holding a pan and is stirring something in it. The kitchen is well-equipped with various appliances and utensils. There is an oven, a sink, and a refrigerator. The woman is also using a knife and a spoon. The kitchen is filled with various items such as bottles, bowls, and a potted plant. The woman appears to be in the process of preparing a meal. The video captures the woman's actions and the surrounding environment in a realistic and detailed manner."
+    ],
+    "video_ids": [
+        "2cvXsi1FUoY_36_0to138"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a pan, a knife, a spoon? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman in a kitchen setting, using a pan on a stove. She is also holding a spoon and appears to be cooking. While a knife is not visibly present in the frames, the core objects mentioned (woman, pan, spoon) are all present and correctly depicted, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a woman in a kitchen,, holding a pan to a stove. She is stirring holding a spoon, appears to be stirring. The there knife is not visible present in the scene provided the presence elements ( ina, pan, spoon) are all present and in identified. fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fcd8a552f91df24184fcbd48706b045c40d1fd06
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45d30838da40814d952b32a77f02ad7464c92c2441d2943f6be85a0d5e57360b
+size 156048
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..53e2899462d11e70bbb9267dd8e4e0a5a7cbcd34
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a blue Mustang car on display at a car show. The car is positioned on a blue floor, and its sleek design and shiny exterior are highlighted. The car's front is prominently featured, with the headlights and grill clearly visible. The car's color and design suggest it's a modern and stylish model. The car show setting implies that the car is a new or special edition model, drawing attention from car enthusiasts and potential buyers. The video captures the essence of the car show experience, where the focus is on the vehicles and their design."
+    ],
+    "video_ids": [
+        "BW6pfUkINOk_1_0to123"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue Mustang car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a blue Mustang car with white racing stripes, showcasing its front grille, headlights, and hood. The car's design and color match the description, and although other vehicles and background elements are visible, they do not contradict the core object being a blue Mustang.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a blue car car, a accents stripes, which its front and, headlights, and part. The car is design and color match the description of indicating there the elements are parts elements are present, they do not detr the main description description a blue Mustang.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1270ac069f440bef2cf689058d1ca5f1f81194cd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52c9e99b04a4285174a7d7a1a59c5aecae3de2729b8c034b84861eb1f3a4bc14
+size 193797
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..34034e9d3f96b138b93f4b1ef3ab6c92aa44e7d3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a serene and peaceful depiction of a small hut in a tropical forest. The hut, with its thatched roof and yellow walls, is surrounded by lush greenery and towering palm trees. The sun casts a warm glow on the scene, highlighting the vibrant colors of the foliage and the rustic charm of the hut. The hut is situated on a dirt path, which leads the viewer's eye deeper into the forest. The overall style of the video is naturalistic, capturing the beauty of the tropical environment in a realistic and unobtrusive manner. The video is likely intended to evoke a sense of tranquility and connection with nature."
+    ],
+    "video_ids": [
+        "4YkyOHQjlLg_8_28to202"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small hut, lush greenery, and towering palm trees.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a small hut with a thatched roof, surrounded by lush greenery and towering palm trees, which aligns perfectly with the described conditions. The presence of a person in a corner does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a small hut with a thatched roof, surrounded by lush greenery, towering palm trees. which aligns with with the ' '. The scene of the sandy walking the yellow of not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2524116de38bdb25c4e9be1ffa207cc2562e1053
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c8f1908f83886b5a82366f91fe4c615dc1718e8a0d22b2d629b568b2618a28b
+size 301301
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..a259a33f3e9101bc44d48c6f81d0298cac337dde
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a kitchen, preparing food. She is wearing a pink shirt and an apron, indicating that she is in the process of cooking or baking. The kitchen is well-equipped with various appliances such as a refrigerator, microwave, and oven. The woman is seen using a spatula to spread something on a tray, which is placed on a marble countertop. The overall style of the video suggests that it is a cooking tutorial or a food preparation video. The woman's actions and the kitchen setting are the main focus of the video."
+    ],
+    "video_ids": [
+        "-laqqQHKOqg_13_31to155"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a spatula, a tray? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman using a spatula to transfer cookies from one tray to another. All three specified objects \u2014 a woman, a spatula, and a tray \u2014 are present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a woman in a spatula to spread or from a tray to another. The the elements objects ( a woman, a spatula, and a tray \u2014 are present and used depicted in the scene.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2e31c7fb62142d94fc0c7ef7adcdfe5cbd2aa14b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aeedee0f21b4f9564424508618ec3d186d3ce32a591ec1adf357b0881ca06472
+size 151331
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..f24445374a94721502ed96964d627777dbbd1cab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a suit holding a whiteboard with percentages and pictures of fruits and vegetables. The man is standing in a room with a window in the background. The whiteboard has the numbers 16% and 3% written on it, along with pictures of a carrot and an apple. The man is pointing to the numbers on the whiteboard. The style of the video is informative and educational."
+    ],
+    "video_ids": [
+        "R32DF4esv4k_4_0to166"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit, a whiteboard, a carrot picture, an apple picture.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit holding a whiteboard with a carrot picture and an apple picture drawn on it, which matches the core description. Additional elements like percentages and grapes are present but do not contradict the required objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit standing a whiteboard. a carrot picture and an apple picture on on it. which align the description description. The elements like the and a are present but do not contradict the main objects.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f93fe5b92445d4b9a69810a1cbd1022cba7f3838
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b4e5c57c3cdab3164f4ae31d2108a2be020facf4622e3ca91d93e4784b22aea2
+size 138595
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..6c592adea2e14afed75fe6cb47c88b9ee72401d6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person holding a bowl of enchiladas, which are a type of Mexican dish. The enchiladas are covered in a sauce and topped with a green garnish. The person is holding the bowl over a table, which has other dishes and utensils on it. The video also includes text that provides the price of the enchiladas in pesos and US dollars, as well as the price of a lemonade. The style of the video is casual and seems to be taken in a restaurant or a home setting. The focus is on the food and the price, rather than the surroundings or the person holding the bowl."
+    ],
+    "video_ids": [
+        "AOF7hqoZD90_28_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a bowl of enchiladas, a table, other dishes and utensils? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person holding a bowl of enchiladas, placed on a table with other dishes and utensils visible in the background. The core elements described are all present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a person holding a bowl of foodiladas, which on a table. other dishes and utensils visible in the background. The presence elements of in present present and accurately with the description content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..48f4e3c602ecaa8430f189c3e644f1a823ba73f3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a1b8eec26299ffdf0e3c2d0d88223d2f691d26e7c7976063dc44e162aff46c6
+size 169221
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..31d0b4765f41ae5f4495028b45d80bb7275c214b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a silver Honda SUV concept car on display at an auto show. The car is positioned on a circular platform, which is surrounded by a metal railing. The car is the main focus of the video, with its sleek design and shiny silver exterior. The background features a large screen displaying the words \"URBAN SUV CONCEPT\" and \"HONDA\", indicating the brand and model of the car. The lighting in the video highlights the car's design and features, making it stand out against the backdrop. The video is likely a promotional or informational piece, aimed at showcasing the car's design and features to potential buyers or enthusiasts."
+    ],
+    "video_ids": [
+        "0OSZ4MrhADE_5_123to251"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Silver Honda SUV concept car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a silver Honda SUV concept car displayed on a rotating platform, consistent with the description. The car's design, color, and branding (Honda logo and 'URBAN SUV CONCEPT' signage) align with the specified object. Additional elements like people and background displays do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a silver Honda SUV concept car, on a rotating platform. which with the description of The car is design, including, and the areHonda logo) 'HondaBAN CON CONCEPT' text) align with the ' object. The elements like the and the details do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f697ca21643efd8e070089b2e072ef13d5ae47eb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1348-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a10c1cd8f597e25df9a1e3ed5a65e71488fc08761eedf110d33669a9b156dc84
+size 150089
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..b3ea63f9e3e7886bd9153895d3d747fbd44d6db3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a dynamic scene on a wooden floor where a large number of colorful dominoes are arranged to form a detailed image of Snoopy, the beloved cartoon dog character from the Peanuts comic strip. The central focus is a grid of white dominoes forming Snoopy's face, with black dominoes outlining his features and adding depth to his expression. Surrounding this central image, a vibrant array of multicolored dominoes creates a border and additional decorative elements.  As the video progresses, the dominoes begin to fall in a chain reaction, starting from the left side of the frame and moving towards the right. The dominoes cascade down in a fluid motion, creating a visually captivating effect. The falling dominoes spread outwards, forming a wave-like pattern that adds a sense of movement and energy to the scene. The camera remains stationary throughout, capturing the entire sequence from a top-down perspective, allowing viewers to observe the intricate details of the domino arrangement"
+    ],
+    "video_ids": [
+        "0e8c046674e4eb5d353e3097704375cc9ad823e61a576424f2ed0cbd689525d5"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dominoes, forming an image of Snoopy, with a central grid of white dominoes and black dominoes outlining Snoopy's features. Multicolored dominoes create a border and additional decorative elements.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large arrangement of dominoes forming an image of Snoopy. The central grid consists of white dominoes with black dominoes outlining Snoopy's features, accurately matching the description. Additionally, multicolored dominoes create a border and decorative elements around the central image, which is consistent with the given condition. The dominoes fall in sequence, but this does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a collection arrangement of dominoes forming an image of Snoopy. The central grid of of white dominoes, black dominoes outlining Snoopy's features, which matching the description. The, multicolored dominoes are a border and other elements around the image image, which is also with the description description. The presenceinoes are slightly a, which this does not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8ba132fb06365fdce01abad1d8f3844d9658b8bb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce524e54d69c39f156c99726d9cb4ad7f7ace0aa7058d44b4f0fcd2f801d08f4
+size 121559
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..75abebb637cabd53140d6581403a86bb2d5d378d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a chocolate-covered dessert, which appears to be a cake or a pie, being prepared. The dessert is covered in a rich, dark chocolate glaze, and it is garnished with several chocolate chips and dollops of whipped cream. The chocolate chips are scattered across the top of the dessert, adding texture and visual interest. The whipped cream is piped in a decorative pattern, enhancing the dessert's presentation. The dessert is placed on a white plate, which contrasts with the dark colors of the chocolate and the cream. The video captures the dessert in three different stages of preparation, showcasing the transformation of the dessert from a simple base to a finished, visually appealing treat. The style of the video is a close-up food preparation video, focusing on the dessert and its garnishes. The camera angle is close, allowing the viewer to see the details of the dessert and its garnishes. The lighting is bright, highlighting the colors and textures of the dessert and its garnishes. The video does not contain any text or narration, allowing the viewer to focus solely on the dessert and its preparation."
+    ],
+    "video_ids": [
+        "nXOOr41DOoc_1_0to170"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chocolate-covered dessert, chocolate chips, whipped cream? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chocolate-covered dessert, which appears to be a cake, generously topped with chocolate drizzle, peanut butter sauce, whipped cream dollops, and peanut butter cup pieces. While it does not explicitly show 'chocolate chips' (which are typically small, distinct pieces of chocolate), the overall composition matches the core description of a chocolate-covered dessert with whipped cream and chocolate elements. The presence of peanut butter cups can be considered a form of chocolate-based confectionery, aligning with the theme.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chocolate-covered dessert with which is to be a cake, with topped with whipped chipsizzle and chocolate butter,, and cream,lops, and chocolate butter cups pieces. The the does not explicitly show chocolatechocolate chips,' aswhich are typically small, round pieces), chocolate), the presence presentation and the description description of a chocolate-covered dessert with whipped cream. chocolate elements. The presence of peanut butter sauce and be considered an variation of chocolate, toppingfectionery, whiching with the theme of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b4d4ece45bc8dccfa781d93bc82043651800bc00
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:817a06961fba82f61b67bd85791061b5b42a36ff5cd0b9fe22dd6f288178c858
+size 88116
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..8c45d976722a5a995ee0a5b149d3b37190e50b8c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a football player in action, wearing a red and white uniform with the number 11. He is holding a football in his right hand, ready to throw it. The player is on the field, surrounded by other players, and the crowd in the background is blurred, indicating the focus is on the player. The style of the video is dynamic and action-packed, capturing the intensity of the game. The player's stance and grip on the football suggest he is preparing to make a play. The video is likely taken from a live game or a highlight reel, showcasing the player's skills and the excitement of the sport."
+    ],
+    "video_ids": [
+        "SfJ4NHJoA-I_1_0to130"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A football player in a red and white uniform with the number 11, holding a football in his right hand, surrounded by other players.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a red and white uniform with the number 11, holding a football in his right hand. Although the background is blurred and other players are not clearly visible, the core description is met. The presence of a blurred crowd or other players in the background does not contradict the main subject's depiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player in a red and white uniform with the number 11, holding a football in his right hand. The the player includes slightly, includes players are present clearly visible, the main elements of met as The player of a blurred background and other players in the background does not contradict the main subject of description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fa9c7596542023788aec85461e4635bea28feb43
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34dc7b8dbf0743df1b2f676f1d14a2dcc1bcbb920242cded5d371eeed6eca916
+size 211951
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..b3b90fe2adaa3507867afa1f910172558c4cc740
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two men are seated at a table, engaged in a conversation. The man on the left is dressed in a black Batman costume, complete with a cowl and cape. His counterpart on the right is wearing a red and blue Robin costume, complete with a mask and gloves. They are seated in a room with a blue wall, and a green chair is visible in the background. The men appear to be in the middle of a discussion, with the man in the Robin costume holding his hands together in a gesture that suggests he is listening intently to the man in the Batman costume. The overall atmosphere of the video is casual and friendly, with the two men appearing to be enjoying their conversation."
+    ],
+    "video_ids": [
+        "O-ZuTFIypQY_46_0to182"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men in costumes (Batman and Robin), one blue chair.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men dressed as Batman and Robin, sitting at a table. While the description mentions 'one blue chair,' the video does not clearly show any blue chair; however, the core elements \u2014 two men in Batman and Robin costumes \u2014 are present and correctly identified. The absence of a visible blue chair does not contradict the main condition, as the presence of additional elements is acceptable as long as they don't conflict with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two characters in in Batman and Robin, seated at a table. There there video mentions aone blue chair,' the video actually not specify specify a blue chair. instead, the presence elements of the men in superhero and Robin costumes \u2014 are present, match depicted. The presence of a blue blue chair does not contradict the main condition, as the focus of the elements ( acceptable as long as they do't conflict with the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d5042b4118d357c8a338307002b42f2580b76ea6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8587b33f341fe859865547b138cac7eb7d083469eca63e26b2b36316ccb99c17
+size 144072
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..ee411dcc36addce8113f3e15527cef1b7435434a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen sitting on a red couch, engaged in a conversation. He is dressed in a black jacket and holds a white cup in his hand. The couch is positioned in front of a red car, which is parked in the background. The man appears to be relaxed and comfortable, suggesting a casual and friendly atmosphere. The red car adds a vibrant touch to the scene, contrasting with the man's black jacket. The overall setting suggests a modern and stylish environment."
+    ],
+    "video_ids": [
+        "GVfZ2mYml9k_106_0to138"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a red couch, a black jacket, a white cup, and a red car.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man sitting on a red couch, wearing a black jacket, holding a white cup, with a red car (Volkswagen) visible in the background. All specified objects are present and accurately represented in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man sitting on a red couch, wearing a black jacket, with a white cup, and a red car visibleaolkswagen) in in the background. The the elements are present and match depicted in the scene.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b0a2aafb6812433761e3762cfb6237e6f06423cd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e85305d13cad39c3c022b1ace513caf209998ac20c6a3ab258abd249c1fb05f
+size 114716
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..903c2da324450c240d23721563077829c7a04009
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a scene from a movie or TV show featuring two women in a conversation. The woman on the left is wearing a blue hijab and a pink blouse with a gold necklace. She is looking at the woman on the right with a surprised expression. The woman on the right is wearing a yellow blouse with a floral pattern and has her hair up. She is speaking to the woman on the left. The background is a blurred image of a building, suggesting an urban setting. The style of the video is realistic, with natural lighting and colors. The focus is on the interaction between the two women, with the background serving as a setting for their conversation."
+    ],
+    "video_ids": [
+        "IN4tMs4inK0_64_138to313"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two women? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two women interacting. One woman is wearing a blue hijab and a pink outfit, while the other is seen from behind wearing a yellow floral top. Both are present and engaged in conversation, fulfilling the 'Two women' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two women standing with One is is wearing a blue hijab and a pink and, while the other is in in the, a yellow floral dress. The are positioned in engaged in the, fulfilling the 'Object women' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..18cb202bb1015b6b5d7ee7d25cc265ecca3974ca
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:20de3d0c19a3fa42de943b2e497014aaa8696a83f3a1e4de24f0afc0686f0a67
+size 144657
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..40c2b65ac7881a0be01c69e154eb08b5071abe32
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a pond teeming with life. The water is crystal clear, revealing a sandy bottom dotted with patches of aquatic plants. These plants, primarily green with some yellowing leaves, create a natural habitat for the fish swimming gracefully through the water. Among them, several koi fish stand out with their vibrant orange and white patterns, adding splashes of color to the tranquil setting. The fish move fluidly, occasionally gliding close to the surface where floating lily pads provide a resting place. The overall atmosphere is peaceful, with the gentle ripples on the water's surface enhancing the sense of calm. The camera remains stationary throughout the sequence, allowing viewers to fully absorb the beauty and tranquility of this aquatic environment."
+    ],
+    "video_ids": [
+        "e014ee9fcb76253459fe70e2d9ea9a42418f6d96144acb140de29e93f1b32bc9"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Fish, koi fish, aquatic plants, lily pads? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows koi fish swimming in clear water, surrounded by aquatic plants and lily pads, which matches the described 'Object(s)' condition. The visual elements are consistent with the requested subjects without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts fishoi fish swimming among a water, surrounded by aquatic plants and lily pads. which are the description 'Object(s)' condition. The presence elements in consistent with the description description, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c802557b58aa661d7501f82fc6b41d79055ef2f2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c92dedd26e6085aed91fb79853a33df9c25a0b4e209960c1106a3186a99a04ea
+size 324690
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..a9502a2fdbe81f25a0240215a00805fcbb0b5b5c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a pink shirt, who appears to be deep in thought. He is seen in three frames, each capturing a different moment of his contemplative expression. The first frame shows him with his hand on his chin, suggesting he is pondering something. In the second frame, he has his hand on his cheek, indicating a moment of reflection. The third frame captures him with his hand on his forehead, possibly indicating a moment of realization or decision-making. The background is blurred, drawing focus to the man and his thoughtful expression. The overall style of the video is a close-up, slow-motion shot, emphasizing the man's facial expressions and the intensity of his thoughts."
+    ],
+    "video_ids": [
+        "AuWLWHHpXks_6_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a pink shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a pink shirt, which matches the 'Object(s)' condition. The man is clearly visible, and his shirt is distinctly pink, fulfilling the requirement without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man wearing a pink shirt in which matches the descriptionObject(s)' condition. The man's consistently the in and the attire is consistently pink. fulfilling the description of any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..edcb0f07bdfc642c0a764f320a448d7c5d267313
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1349-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:495b84ea7354497067dbd2f558d0c68851ff2e1b55c0fbc72ea25ee7adf9d96a
+size 125288
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..17ae7b8570ef0a8b01c1dd7453f3778f055021ee
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video opens with a serene view of a vast expanse of blue sky, which gradually transitions into a tranquil ocean below. The sky is mostly clear with a few soft, wispy clouds scattered across it, suggesting a calm, sunny day. As the camera pans downward, it reveals the expansive surface of the ocean, its waves gently rippling under the sunlight. The water appears deep blue, with subtle variations in shade indicating depth and movement.  In the foreground, a bird, likely a seagull or a similar seabird, is captured in flight. Its wings are spread wide, showcasing the intricate patterns on its feathers, which are predominantly dark with lighter edges. The bird's body is streamlined, and its tail is slightly forked, typical of many seabirds. It soars gracefully, moving from left to right across the frame, adding a dynamic element to the otherwise still seascape. The bird's flight path suggests it might be searching for food or simply enjoying the freedom of the"
+    ],
+    "video_ids": [
+        "0cfd22ba00c0f7571a7825cec2853206a29167b65a3f27998d9b0c7f38ef53a3"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bird, likely a seagull, flying in the scene.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bird in flight over a body of water, which matches the description of a bird, likely a seagull, flying. The bird's movements and environment are consistent with the given condition, even though the bird's exact species is not confirmed as a seagull, it fits the general description of a bird flying in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bird in flight over a body of water, which align the description of a bird, likely a seagull, flying in The bird is wings and the are consistent with the given condition.\"\n and though the bird's specific species is not specified. a seagull. it is the general description of a bird flying.\"\n the scene.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f1d7e881641dc053070a2d09bca1042d94795769
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:15a5f5cfd4fdd33f7270a4aeac7fa965ea7cba79deedd8e6db5c22f95e8c7169
+size 191982
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..9945b6774cdb45485e7ca4b45b7060d893c9325e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the vibrant beauty of a pink flower bush in bloom. The flowers, in various stages of opening, display a range of pink hues, from soft blush to deeper magenta. The bush is lush with green leaves that provide a striking contrast to the flowers. The flowers are arranged in clusters, creating a dense canopy of color. The video is a close-up shot, allowing the viewer to appreciate the intricate details of the flowers and leaves. The style of the video is realistic, capturing the natural beauty of the flower bush in its full glory."
+    ],
+    "video_ids": [
+        "pWGb4hGjAtI_32_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Pink flowers in various stages of opening, green leaves? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows pink flowers in various stages of opening, surrounded by green leaves, which matches the described 'Object(s)' condition. The flowers are vibrant and appear in different states \u2014 some fully bloomed, some partially open, and some wilting \u2014 while the green leaves provide a natural backdrop. No conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a flowers in various stages of opening, which by green leaves, which align the description 'Object(s)' condition. The flowers are depicted and the to different stages of some fully bloomed, others partially open, and some stillting, indicating the green leaves are a natural backdrop. There additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..90bf6817b7df3f7e46da9ad41f497340bf404f5a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03d81677ac3f21301c4370f995e3b7f184a3cc13b677d70d28afe8492ee321be
+size 119083
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..477392ded4d0f20f4ab95605d9f3f4452642f891
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a blue shirt with a mustache, engaged in a conversation with another person whose face is not visible. The man appears to be speaking, with his mouth slightly open and his eyes focused on the other person. The setting appears to be an indoor room with warm lighting, possibly a living room or a similar domestic space. The style of the video is a close-up shot, focusing on the man's face and upper body, which suggests a personal or intimate conversation. The image quality is clear, with good lighting and color saturation, indicating that the video is likely professionally shot or edited."
+    ],
+    "video_ids": [
+        "JlVYr3olXqE_606_0to186"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue shirt with a mustache, another person (face not visible).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a blue shirt with a mustache, and another person whose face is not visible, matching the description. The background and lighting are consistent with a scene from a film, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue shirt with a mustache, and there person is face is not visible. as the description provided The man and lighting are consistent with the typical that a video or and there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8f88a857c9b52d11b4f5677170c559ec3cdf8267
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6f5c65ecb88c89ebd92b69d047c03e60c22f3122456ef2c5f4d2316d37726aab
+size 146896
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..28671bf6d59fa65c5d08e9c6df36717628f940eb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the growth of a field of tall grasses and plants, starting from a close-up view of the green shoots and ending with a wide shot of the fully grown plants. The plants start as small, green shoots and gradually grow into tall, brown stalks with green leaves. The field is lush and vibrant, with the plants reaching up to the sky. The video is shot in a close-up style, focusing on the details of the plants and their growth. The colors in the video are predominantly green and brown, reflecting the natural hues of the plants. The video is a testament to the beauty of nature and the power of growth."
+    ],
+    "video_ids": [
+        "9GU0q0zGQIc_22_0to104"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Tall grasses and plants? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a field of tall grasses and plants, specifically what appears to be sorghum, with brownish seed heads and green leaves, which matches the description of 'Tall grasses and plants'. There are no conflicting elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a close of tall grasses and plants, with focusing appears to be aghum or which a and seed heads and green leaves. which align the description of tallTall grasses and plants'. The are no additional elements in contradict this description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2d5b3ec00746bb99afda5c36194cddf7f7572d68
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b84d5c106a8a5996208f7511963c1927be54deb1f674eac13787d282948433c8
+size 240319
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..944bd87219b45d9fa68ba0160604cccdda8ca36c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a classroom setting with a woman teacher standing at a desk. She is wearing a plaid shirt and has her hands clasped together. On the desk, there are three dolls, each with a different color dress: yellow, blue, and pink. The classroom has a colorful money chart on the wall, and there are windows with blinds. The style of the video is educational, with a focus on teaching children about money. The teacher appears to be engaging with the dolls as part of her lesson."
+    ],
+    "video_ids": [
+        "G3c-N-gL0zs_10_0to153"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Woman teacher, three dolls, desk? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman who appears to be a teacher sitting at a desk. There are three dolls positioned on the desk in front of her. The setting and objects match the described condition, even though there are additional classroom posters on the wall, which do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a woman in appears to be a teacher, at a desk with There are three dolls positioned in the desk in front of her, The setting and the in the description ' without with though the are additional elements elements and the wall in which do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..904c6ab5be68f2067b696e77901ad9de0a414fd5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c9720fdfce5cb5887bfb7c92c26bc31de2db1d611f98821a705588cde947d487
+size 189770
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..e9cb881bc3fd5ed66f0ea2fee7cd3bdd26501568
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of joy and celebration, featuring a man and a woman standing together in a crowd. The man, dressed in a suit and tie, is waving to the crowd with a warm smile on his face. The woman, wearing a yellow jacket, is standing next to him, also smiling and waving. They are surrounded by a group of people, all of whom appear to be in a good mood. The crowd is diverse, with people of various ages and backgrounds. The setting appears to be an outdoor event, possibly a political rally or a public gathering. The atmosphere is lively and energetic, with the crowd cheering and applauding. The man and woman seem to be the main focus of the event, drawing the attention of the crowd. The video is a snapshot of a moment of unity and celebration, capturing the spirit of the event and the enthusiasm of the crowd."
+    ],
+    "video_ids": [
+        "OMA7-h8kCAE_2_94to229"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman standing together in a crowd.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a woman standing together in a crowd, which matches the 'Object(s)' condition. The man is Barack Obama, and the woman is Michelle Obama, both prominently featured in the foreground. The background contains other people, forming a crowd, which aligns with the description. There are no elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a woman standing together in a crowd. which align the 'Object(s)' condition. The woman is wearing Obama and and the woman is Michelle Obama, as of featured in the crowd. The crowd is a people, indicating a crowd, which iss with the description. The are no additional in contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d3a2fb538ab40752c3334fe801aa986f8633c154
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd57f6e0bb7580da06f240ef5963d39390ecf3c5ec7b73781284063c2e7365b7
+size 297503
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..2cb4b0c0e0d466365cbf42ec5b6277078cc2e8e2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a bald man wearing glasses and a pink visor, sitting in front of a microphone. He is dressed in a maroon robe and has his hands clasped together in a gesture of prayer or contemplation. The setting appears to be a stage or a public speaking event, with a backdrop of plants and a wooden structure. The man's expression is serious and focused, suggesting that he is delivering a speech or a lecture. The style of the video is straightforward and documentary-like, capturing the man's actions and the setting without any additional embellishments or artistic effects."
+    ],
+    "video_ids": [
+        "CJl0KsMHtYo_57_0to120"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bald man wearing glasses and a pink visor, dressed in a maroon robe, with his hands clasped together.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man wearing glasses and a pink visor, dressed in a maroon robe, with his hands clasped together. These elements are consistent with the description, and no significant contradictions are present. Additional elements (like the microphone, background plants, or another person partially visible) do not conflict with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man wearing glasses and a pink visor, dressed in a maroon robe, with his hands clasped together. The elements match consistent with the description provided and there additional contradictions are present.\"\n The elements suchsuch the microphone and background plants, and the person's visible) do not conflict with the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4da65ded97bb7cd3f04166270857bd937e5ce4a8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f9eb19ecd4a44eb735a382c89033074ec2931d0209f281d85f8d9775f4f6b7fd
+size 157076
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..dee23a04350a3fd89957603362de836775e3aa4e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a glass bowl filled with a colorful pasta salad. The salad consists of small pieces of pasta, green peas, orange carrots, and red bell peppers. The ingredients are mixed together, creating a vibrant and appetizing display. The bowl is placed on a wooden table, and the background is blurred, focusing the viewer's attention on the pasta salad. The style of the video is simple and straightforward, with a focus on the food. The lighting is bright, highlighting the colors of the ingredients and making the salad look fresh and inviting. The video does not contain any text or additional elements, keeping the focus solely on the pasta salad."
+    ],
+    "video_ids": [
+        "vqM4PzxAQ38_4_0to128"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A glass bowl filled with a colorful pasta salad containing small pieces of pasta, green peas, orange carrots, and red bell peppers.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a clear glass bowl containing bow-tie pasta, green peas, orange carrot pieces, and red bell pepper pieces, which matches the description of a colorful pasta salad. The composition and colors align with the specified ingredients, and there are no conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a glass glass bowl filled a-tie pasta, green peas, orange carrot pieces, and red bell pepper pieces, which matches the description of a colorful pasta salad. The presence and elements are with the provided ingredients, fulfilling there are no additional elements in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f838a6d7168393932d5a2929f614ef5148f8c3bd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-135-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c187037ba3b63391c4ebaea4f1827f8531e4637f6294b5c897171a1156c09d2e
+size 97188
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..f545c6f6b2bdc015507b313c32a8c4297f84504a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman with blonde hair is seen in a room with a blue door. She is wearing a yellow sweater and has her mouth open as if she is speaking or reacting to something. The room has a clock on the wall and a coat hanging on the door. The woman appears to be the main subject of the video, and her actions suggest a conversation or interaction with someone off-camera. The style of the video is realistic, capturing a candid moment in the woman's life."
+    ],
+    "video_ids": [
+        "6a1gpbDjYAc_0_0to156"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with blonde hair, wearing a yellow sweater. A clock, a coat.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a woman with blonde hair wearing a yellow sweater, which matches the core description. While a coat is visible in the background (likely hanging), it does not contradict the description. The presence of a clock is not clearly visible or identifiable in the video frames provided, but since the core description is largely fulfilled, the answer is 'Yes'.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a woman with blonde hair wearing a yellow sweater. which matches the description description. Additionally there clock is not in the background,h hanging on it is not contradict the description as The presence of a clock on also mentioned visible in mentioned in the video, provided, but it the description elements does not met, the answer is 'Yes'.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c04a651e14aef5b5a2bfbe63cab36dccec5adb7b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e23317127223a9a771746f46c1019b73d9fe9d12ed358328867c7d4737127e68
+size 244457
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..fab798ddae8cb8132d3e3f5fcfeb3dcc91506341
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a classic car, showcasing its vintage design and features. The car's dashboard is adorned with a variety of gauges and dials, each serving a specific purpose in monitoring the vehicle's performance. The steering wheel, a prominent feature in the center of the dashboard, is made of wood and metal, adding a touch of elegance and sophistication to the overall design. The car's interior is bathed in sunlight, highlighting the details of the dashboard and steering wheel. The video is a testament to the craftsmanship and attention to detail that went into designing this classic car."
+    ],
+    "video_ids": [
+        "x7skkkY-xq4_50_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dashboard gauges and dials, steering wheel? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the dashboard gauges and dials, along with the steering wheel, which are the primary objects mentioned in the condition. The camera focuses on these elements, and they are prominently displayed throughout the video, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a dashboard gauges and dials, as with the steering wheel, which are the main elements mentioned in the description. The image focuses on these elements, providing there are prominently displayed, the video. fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cadbe21c76e81d5deb59236f0891f7ce4f9c98d6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b31cf9fc5dc685264e5429a5b86d453a068d2e48492a9979a2a2466eeee0614
+size 188441
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..25bda161803d4e99cf9b6ef0848aa8f1a7695210
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is on a player in a blue and white uniform, who is in possession of the football. He is running towards the right side of the frame, with his body leaning forward in a sprint. His arms are outstretched, and his eyes are focused on the path ahead.  In the background, two other players are visible. One is in a black uniform, and the other is in a white uniform. They are both in pursuit of the player with the ball, their bodies angled towards him as they try to catch up. The player in the black uniform is closer to the camera, while the player in the white uniform is further away.  The setting is a football field, with the players positioned in the middle of the field. The field is marked with white lines, and the goalposts are visible in the distance. The lighting suggests that the game is taking place during the day.  The style of the video is realistic, capturing the intensity and action of the game. The focus is on the players and their movements, with the background serving to provide context for the action. The video does not contain any text or additional graphics."
+    ],
+    "video_ids": [
+        "2AXI6WmQztU_104_0to123"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a blue and white uniform with a football, a player in a black uniform, and a player in a white uniform.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a blue and white uniform holding a football, a player in a black uniform, and a player in a white uniform, which matches the described 'Object(s)' condition. The presence of additional players and background elements does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a player in a blue and white uniform with a football, and player in a white uniform, and a player in a white uniform. which matches the description 'Object(s)' condition. The presence of the elements and the elements does not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..66b7f8887e1eea69353a6658a2fdc0c1c1aa3c18
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:12c709bf9dab88d6e9640c7af0779b452edf1fd1c38381c76fbe1aba7c9f12db
+size 237686
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..1d55bb7ed1829d80d6d3b6547bf22d6b3476ef6e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of the engine compartment of a car, with the hood open. The engine is visible, with various components and parts, including the air filter, battery, and fluid reservoirs. The engine cover is black with a red and white label on it. The style of the video is a straightforward, unembellished documentation of the car's engine compartment, likely intended for educational or informational purposes. The focus is on the engine and its components, with no additional context or background provided. The lighting is bright, and the camera angle is from above, providing a clear view of the engine and its parts."
+    ],
+    "video_ids": [
+        "_Wcfcijwcd8_14_0to157"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Engine, air filter, battery, fluid reservoirs, engine cover.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows the engine bay of a car with the hood open, revealing the engine cover prominently in the center. The engine cover has a visible logo or emblem. To the left, a fluid reservoir (likely coolant or windshield washer fluid) is visible, and to the right, components that resemble the battery and other engine-related parts are present. The video successfully fulfills the 'Object(s)' condition as it includes the engine, engine cover, and fluid reservoirs as described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a engine compartment of a car, various engine open, revealing the engine cover, in the center. There engine cover is a logo logo, emblem, There the right of there part reservoir islikely the) oil washer fluid) is visible, and to the right, there that could an air and air fluid-related parts are present. The air does fulfills the 'Object(s)' condition by it displays the engine, engine cover, and fluid reservoirs, described.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2938662fd15c6a6f3f2e2d4d18df48ab171e498d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:05d391bb6b1aad4532d516934bed1de558cb40f857856e10597f98e4824d0db4
+size 125299
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..f8aa2c650ab24791356d01e0ab2cc0e9a804881f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a black Mercedes-Benz sports car driving down a street. The car is sleek and shiny, with a large grille and the Mercedes-Benz logo prominently displayed. The car's headlights are on, and it appears to be moving at a moderate speed. The street is lined with buildings, and the car is driving on the right side of the road. The sky is clear and blue, suggesting it's a sunny day. The car's occupants are not visible in the video. The style of the video is realistic, with no visible special effects or animation. The focus is on the car and its movement, with the surrounding environment serving as a backdrop."
+    ],
+    "video_ids": [
+        "B5-5hP2NOdE_11_0to112"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black Mercedes-Benz sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a black Mercedes-Benz sports car, specifically a SLS AMG model, which matches the description. The car is shown in motion and stationary, with clear visibility of its design, logo, and color. There are no elements that contradict the core description of the object.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features features a black Mercedes-Benz sports car, which a model-Class AMG model, as matches the description of The car is shown in motion, the positions with clear visibility of its design and including, and overall. The are no elements in contradict the description description of the object.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1a3ad38a1f217c2249c5f1fa4d3829dd088d4e5b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8d7c0e407166ab56240fd733c6401f50ec4033da531b5ae014c3424402f24bbe
+size 350891
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..c273ad12ed6c6a9661419db7cb0fbe1c8232c48a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a polar bear in an enclosure, partially submerged in water. The bear is seen interacting with its environment, moving its head and body slightly, and occasionally splashing water with its paws. The enclosure features a glass barrier through which the bear is visible, with water droplets forming on the glass, indicating recent movement or interaction with the water. In the background, there are rocks and some greenery, suggesting a naturalistic setting designed to mimic the bear's habitat. The lighting appears to be natural daylight, enhancing the visibility of the bear's white fur and the blue tones of the water. The bear's actions are calm and deliberate, showcasing its natural behavior in a controlled environment."
+    ],
+    "video_ids": [
+        "c1048204197074ce0b71ae4ced4a7687246a52b3e86df6b41348d44e221a016d"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A polar bear, water, rocks, greenery, glass barrier? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a polar bear in water, with rocks and greenery visible in the background, and a glass barrier separating the bear from the viewer. All elements mentioned in the condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a polar bear in water, with rocks and greenery in in the background. which there glass barrier is the bear from the viewer. The elements mentioned in the ' are present and do depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..acb6deea5457447a8c16c3e51ce1f7f249c0392b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8fb973d52b082b48e068acbe60d1d1e977b1a92cb83af3585e790da79994c0a4
+size 229108
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..aec93e8f2d37bba1119a89be993ce276cce57f91
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen in a kitchen, holding a blender filled with a red liquid. He is wearing a green t-shirt and appears to be surprised or excited about the contents of the blender. The kitchen has wooden cabinets and a window, and there is a sink visible in the background. The man's facial expression and the context suggest that he might be trying out a new recipe or experiencing an unexpected result from his concoction. The overall style of the video seems to be casual and informal, capturing a moment of surprise or discovery in a home kitchen setting."
+    ],
+    "video_ids": [
+        "FfbaUOoWRlQ_19_0to213"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a blender filled with red liquid, a green t-shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a green t-shirt holding a drink with a slice of red velvet cake on top. In the background, there is a blender filled with red liquid, matching the described objects. No conflicting elements are present that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a green t-shirt holding a blender in a blender of lime fruit cake in top, The the background, there is a kitchen filled with red liquid, which the description elements. The additional elements are present, would the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..66a190780d0c7b479945c1e77a786a2d42b49545
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:38180bb2fe7bcf45a700fa40f18fd0d09670af07bd2ea38c0af89e8d8b3b2143
+size 154492
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..deddaf3ce23b1059b81fb4bc6e76d822e0250d90
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man sitting in the back seat of a luxury car, using a tablet. The car is well-appointed with leather seats and a sunroof. The man is dressed in a suit and tie, suggesting a professional setting. The car is parked outside a building with a brick facade, indicating an urban environment. The man appears to be engaged in a conversation or presentation, as he gestures with his hands while using the tablet. The overall style of the video is sleek and professional, with a focus on the man and his interaction with the technology."
+    ],
+    "video_ids": [
+        "D1aVDQL6DBM_27_0to120"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a luxury car, a tablet, and a building with a brick facade.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a luxury car, holding a tablet, and a building with a brick facade visible through the window. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit car, holding a tablet, and there building with a brick facade is through the car. The the elements of in present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..784a2d21b2ee0d8698ee2687c475643995c7ea0f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1350-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8df349afef83031f6c3b924b805f7f3c389727f64e8c0e1ad18f070d352e228f
+size 136222
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..4bec0871eabd4d2aa5ca0f90129e09e3c4f24d85
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a blue shirt, captured in a close-up shot. The man is looking directly at the camera, his expression serious. The background is a stark black, which contrasts with the man's blue shirt and highlights his features. The lighting is soft, casting gentle shadows on the man's face. The overall style of the video is minimalist and focused, with the man being the sole subject. The video does not contain any other objects or elements."
+    ],
+    "video_ids": [
+        "IUuQupN4shs_2_72to264"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue shirt against a black background. The man's attire and the color of the shirt match the description, and there are no conflicting elements that contradict the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person wearing a blue shirt, a dark background. The description is attire and the presence of his shirt match the description provided fulfilling there are no additional elements in would the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6ebebfbcad0e2be845c54ecbbce634b1cb1a9deb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f819676ab89983b433c444f7220fa63aa2c031970782d99df07d0af84ca237de
+size 81998
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..17583e0e5245224809e068fde39df6621210bf2f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a luxury car, showcasing its design and features. The car's dashboard is adorned with a sleek, black and silver color scheme, featuring a large touch screen display at the center. The steering wheel, also black, is equipped with multiple buttons and a logo on the center. The car's interior is further accentuated with a wooden panel on the dashboard, adding a touch of elegance. The video is a close-up shot, focusing on the car's interior, and is likely taken from the perspective of the driver. The style of the video is that of a promotional or advertisement material, highlighting the car's luxurious features and design."
+    ],
+    "video_ids": [
+        "Ya71zzlcdnM_20_0to181"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dashboard (sleek, black and silver), large touch screen display, steering wheel (black), multiple buttons, logo? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a sleek dashboard with black and silver accents, a large touch screen display, a black steering wheel with a visible logo, and multiple buttons and controls. All elements described in the condition are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a sleek dashboard with a and silver elements, a large touch screen display, a black steering wheel, multiple logo logo, and multiple buttons on controls. The these match in the condition are present and match with the image content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ec92e2d029ce83ff70f021239d76177aa2e86d3c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52af409966d2360d002572821d4284de6136a6e5887db3f17991ce007a4c70c4
+size 161966
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..87b2d07b73b9576a2ada71cc44c68bc59ca57882
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit sitting at a desk, with a blurred background of a room with a lamp. The style of the video is artistic, with a focus on the man's hands and the desk in front of him. The background is blurred with blue and yellow wavy lines, creating a sense of motion and energy. The man appears to be focused on something on the desk, possibly a document or a computer screen. The overall mood of the video is calm and contemplative, with a touch of creativity and imagination."
+    ],
+    "video_ids": [
+        "GIyuXxs6KMw_102_0to197"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit, his hands, a desk, a document or computer screen? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a suit sitting at a desk, with his hands visible on a document. The setting includes a desk and a document, which aligns with the 'Object(s)' condition. Although there is a lamp and other office items, they do not contradict the core description. The visual elements are consistent with the requested objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a man in a suit, at a desk. with his hands positioned as a document. The presence includes a desk and a document, which aligns with the 'Object(s)' condition. The there are a colorful and some elements elements in they do not contradict the core description and The presence elements are consistent with the description objects.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dbb4f89af3049b4259ba329dccc8001ff1a9382a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41ba16ed32c3a3057a19d2927fbfb88e4e13f62b324d226588ba000acb40e7dc
+size 81225
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..d665e4ff06e759d65d0c7bf5e61eaa67e72ae40a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a luxurious speedboat cruising on a large body of water. The boat, painted in white and black, is sleek and modern, with a curved windshield and a cabin that houses the driver and a passenger. The boat is moving at a moderate speed, leaving a trail of ripples in its wake. The water around the boat is calm, reflecting the clear blue sky above. In the distance, a city skyline can be seen, adding a touch of urbanity to the otherwise natural setting. The boat's journey across the water is the main focus of the video, with no other significant actions or interactions taking place. The overall style of the video is serene and peaceful, capturing the beauty of a leisurely boat ride on a sunny day."
+    ],
+    "video_ids": [
+        "yJfieAY19Ac_16_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A luxurious speedboat, a city skyline? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a luxurious speedboat named 'Fata Morgana' cruising on water, with a city skyline visible in the background. The boat's design, including its sleek lines and modern features, matches the description of a luxurious speedboat. The city skyline with tall buildings is also prominently displayed, fulfilling the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a luxurious speedboat, 'Serleana' on on a. which a city skyline visible in the background. The presence is design and with its sleek white and modern features, align the description of a luxurious speedboat. The presence skyline in tall buildings further also present visible, fulfilling the condition of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..910cb19696bd65f0a4c82519c7c36500d58bd571
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b89bfb53a1896bb517375a8452abb5b54dffe5df0db8733bd0a6131215cce24f
+size 252395
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..8d455dcbd582be2b61d3210f9e0b38447236ee0e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man with a beard and glasses, wearing a blue shirt, taking a selfie in an office environment. The office is filled with various objects such as shelves, cabinets, and a desk with a computer monitor. The man appears to be in a good mood, as he is smiling and looking directly at the camera. The office has a professional and organized atmosphere, with papers neatly arranged on the desk and shelves. The lighting in the office is bright, and the overall ambiance is calm and focused."
+    ],
+    "video_ids": [
+        "cMNh_cioHj8_0_0to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man with a beard and glasses, wearing a blue shirt, taking a selfie.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and glasses, wearing a blue patterned shirt, holding the camera at arm's length as if taking a selfie. The background is an office environment, which does not contradict the description. All core elements of the description are present and consistent.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and glasses, wearing a blue shirted shirt, taking a camera to arm's length to if taking a selfie. The background appears an office setting with which is not contradict the description. The elements elements of the description are present in accurately with}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cb1730ddfccdb03e8f2b63918eecb42dac104004
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04d71625268a06b35c089c513b74b15538d29deebc07df2c19b670d739849edd
+size 171824
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..80607f7251c5e2f30983e4c5e245350b135bf0a2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men in a workshop setting, engaged in a conversation. The man on the left is wearing a gray sweater and glasses, while the man on the right is dressed in a black shirt. They are standing in front of a workbench, which is equipped with various tools and equipment. The workshop appears to be well-lit and organized, with a blue machine visible in the background. The men seem to be discussing something, possibly related to their work or a project they are working on. The overall style of the video is candid and informal, capturing a moment of interaction between the two men in their professional environment."
+    ],
+    "video_ids": [
+        "Bg2rGEVLoTY_81_0to117"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, a workbench, various tools and equipment, a blue machine? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men standing and talking in what appears to be a workshop or lab setting. A workbench is visible behind them, and there are various tools and equipment on it. A blue machine (likely a paper towel dispenser or similar device) is mounted on the wall. All elements described in the condition are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men standing in talking in front appears to be a workshop or lab setting. There workbench is visible in them, with various are various tools and equipment on the, Additionally blue machine ispossibly a machine cutter dispenser) similar)) is also on the wall in The elements in in the condition are present in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b72fcf22c63e33dc15cd14f6c414250e8e08bbc3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1cb91113aebc19ffebb1d18ff4aa00a8eae8c7261b723bb0f92c17b1c7101e4a
+size 143265
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..eb585acefe9f93aa1c8daaf894998519f26b7d56
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a brown and white guinea pig with curly fur. The guinea pig is sitting on a brown blanket with a blue polka dot pattern. The guinea pig is looking directly at the camera with its eyes wide open. The guinea pig appears to be calm and comfortable in its environment. The video is likely a close-up shot of the guinea pig, focusing on its fur and facial features. The style of the video is likely to be a simple, straightforward shot with no additional elements or actions."
+    ],
+    "video_ids": [
+        "EqTx9qzuSU0_18_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A brown and white guinea pig with curly fur.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a guinea pig with brown and white fur and curly hair, which matches the description. The guinea pig is resting in a cozy setting, and while there are additional elements like a blanket and a toy, they do not contradict the core description of the animal itself.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a guinea pig with brown and white fur, curly hair, which matches the description. The presenceinea pig is the on a cozy, with and the there are additional elements like the blue with text text, they do not contradict the core description of the gu.\"\n.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..05fe9e62a188d3cd12b8d568ac993e2db86e57cc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30f54eac348ef43bf9fa16c3f69ee466b29d4a707844490a847a1b548cf025fd
+size 187508
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..214680ecb32d33ca45279f258a9d35298a6f4019
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 4K resolution film demo, showcasing a woman in a white ball gown with a long train, standing in a luxurious room with red walls and gold trim. She is positioned in front of a large window with a chandelier hanging from the ceiling. The room is adorned with a large rug and a grandfather clock in the background. The woman is looking over her shoulder, giving a sense of elegance and sophistication. The overall style of the video is classic and refined, with a focus on the woman's attire and the opulence of the room."
+    ],
+    "video_ids": [
+        "8ygHYbtZK3s_76_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a white ball gown with a long train, a chandelier, a large rug, and a grandfather clock.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman in a white ball gown with a long train, a chandelier in the foreground, a large ornate rug on the floor, and a grandfather clock in the background. All core elements described are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a woman in a white ball gown with a long train, standing chandelier, the background, a large rugate rug, the floor, and a grandfather clock on the background. The elements elements of in present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5f6969f0c9350fcfdad550e1710e46d431b2af41
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1351-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f4669950cc9a151655b6af881f601a62a7acbbefbd1f9336db3369e79b96e15
+size 186401
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..aa139ae68197e6f8f9ac065ea6f1f5c9bc73f347
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is on two players: a quarterback from the Dallas Cowboys, who is in the process of throwing the ball, and a defensive end from the Detroit Lions, who is attempting to block the throw. The quarterback is in the center of the frame, with his arm extended and the football visible in his hand. He is wearing a white jersey with the number 9 and the Cowboys' logo on the front. The defensive end is positioned to the right of the quarterback, with his arms outstretched and his body leaning forward in an attempt to disrupt the throw. He is wearing a blue jersey with the number 94 and the Lions' logo on the front. The background of the image shows a crowd of spectators in the stands, adding to the atmosphere of the game. The style of the video is realistic, capturing the intensity and action of the game."
+    ],
+    "video_ids": [
+        "KT77QbNz6r4_2_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Quarterback (Dallas Cowboys) and Defensive End (Detroit Lions).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a quarterback in a Dallas Cowboys uniform (number 9) throwing the ball, and a defensive end in a Detroit Lions uniform (number 94) attempting to tackle him. The players' uniforms, numbers, and team affiliations match the description, and the action depicted is consistent with an NFL game scenario.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts two football in a white Cowboys uniform (white 9) and a ball, and a defensive end in a Detroit Lions uniform (number 39) attempting to block him. The presence' uniforms and including, and the colorsations match the description provided fulfilling the action align align consistent with a NFL game scenario.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f3e7ed24c6563c2b45121dc31d22f66dd63a451b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b7eb407efd53e7b8f99edaa5d975cc4b180a993087cc336b9e1c4d59954e562d
+size 284016
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..b20ca9668898b7547385c52c57cb3b8deb781e52
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a black beanie with the word \"NEW\" on it, standing in front of a building with a metal door. He is wearing a brown jacket and has a mustache. The man appears to be speaking, as suggested by his open mouth and the context of the image. The video is likely a news segment, as indicated by the logo in the bottom right corner, which reads \"5:01 35\u00b0 #abc7ny\". The style of the video is a standard news report, with a focus on the man and his surroundings. The image captures a moment in time, with the man as the central figure, and the building and metal door serving as the backdrop. The overall tone of the video is serious, suggesting that the man may be discussing an important topic or event."
+    ],
+    "video_ids": [
+        "2X2UwhPgJmQ_2_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black beanie with 'NEW' on it, wearing a brown jacket and a mustache. There is also a building with a metal door in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black beanie with 'NEW' visible on it, a brown jacket, and a mustache, standing in front of a building with a metal door in the background. The visual details match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black beanie with 'NEW' written on it, a brown jacket, and a mustache. which in front of a building with a metal door. the background. The description elements match the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e082f7bc40f1b26c1e37b4c16abec4d27ea9adde
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4aed75edcbc03cd7424d7974e371afb72168b57d71970b422b3753ba7614930
+size 131280
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..a6c57774e1e0b66b911bc7a1a6282fef804aa54b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman with long brown hair is seen in a kitchen setting. She is wearing a yellow shirt and is smiling at the camera. The kitchen has wooden cabinets and a sink. On the counter, there is a blender and a bottle of dish soap. The woman appears to be in a good mood and is likely preparing to cook or clean. The overall style of the video is casual and homey, with a focus on the woman and her surroundings."
+    ],
+    "video_ids": [
+        "GLiXsk-kkI0_6_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with long brown hair, wearing a yellow shirt, a blender, and a bottle of dish soap.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with long brown hair wearing a yellow shirt, standing in a kitchen. A blender is visible on the counter behind her, and a bottle of dish soap is present near the sink. These elements match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with long brown hair wearing a yellow shirt. standing in a kitchen with There blender is placed on the counter in her, and there bottle of dish soap is on on the sink. The elements match the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8017cebc46a36a53c1896d625389ec5e5c07794a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:456f30cf87a69be2ed4c17330eaf44b9e8b4745395b4a8e08b9b46825eff9987
+size 189106
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..17e7064c7cdf4d6a6a33daf15c421ac31535ce0a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a vibrant and colorful display of a healthy smoothie being prepared. The first frame shows a glass of red smoothie with a few fresh ingredients sprinkled on top, including a sprig of dill and a slice of apple. The second frame shows the same glass of smoothie, but now with a few more ingredients added, including a slice of beet and a sprig of parsley. The third frame shows the final product, a fully prepared and garnished glass of red smoothie, ready to be enjoyed. The style of the video is simple and straightforward, focusing on the preparation of the smoothie and the fresh ingredients used. The colors are bright and appealing, emphasizing the healthiness and freshness of the ingredients. The video is likely intended to inspire viewers to make their own healthy smoothies at home."
+    ],
+    "video_ids": [
+        "ZwryouK_dfk_46_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ['glass of red smoothie', 'sprig of dill', 'slice of apple', 'slice of beet', 'sprig of parsley']? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a glass of red smoothie topped with a sprig of dill, a slice of apple, and slices of beet. There are also sprigs of what appears to be parsley or a similar herb scattered around the scene. All specified objects are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a glass of red smoothie, with a sprig of parsleyill and a slice of apple, and a of what. The is also sprigs of parsley appears to be parsley. a similar herb. around the glass. The the objects are present, correctly with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..775d91b130c3ba2cd21bebbb1c3f8b57c5bf1457
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb7931e3ea8832f49937f599b269cec740862146d6c2bbc898be4f4cd1442dfc
+size 108048
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..f6adcd0fc78eb7fe8f289448545eb978a84b3c0d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features an older man with white hair, wearing glasses, a suit, and a tie. He is seated in front of a colorful background that includes a cityscape and a sunset. The man appears to be speaking or being interviewed, as suggested by his attentive expression and the presence of a microphone. The video is likely a news segment or an interview, given the professional attire of the man and the setting. The style of the video is formal and professional, with a focus on the man and his surroundings."
+    ],
+    "video_ids": [
+        "dmYgkgWciYI_3_0to189"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An older man with white hair, wearing glasses, a suit, and a tie. A microphone is also present.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an older man with white hair, wearing glasses, a suit, and a tie, which matches the description. A microphone is also visible on his suit, confirming its presence. The video does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features an older man with white hair, wearing glasses, a suit, and a tie. which matches the description. Additionally microphone is also visible on the left lap indicating the presence. The background largely not contradict the given description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ce205477b809f777b87900c55671f467b0a306ba
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8013dda0310f4ed0d0965151b6fbbf3dede470d65fad9dddb9a73eb1795cea0e
+size 216476
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..968a16648df5d480cb03b05687dca1c39b8d871b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seated at a table, engaged in a conversation. He is dressed in a white shirt and has a beard. The table is set with a plate of food, a cup, and a bottle. The setting appears to be a restaurant or a cafe, with potted plants and a brick wall in the background. The man is gesturing with his hands, indicating that he is actively participating in the conversation. The overall atmosphere of the video is casual and relaxed."
+    ],
+    "video_ids": [
+        "BIotSabF_BQ_5_0to120"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a table, a plate of food, a cup, and a bottle.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man sitting at a table with a plate of food, a cup, and a bottle in front of him. These core objects are present and consistent with the description, even though there are additional elements like plants and other furniture in the background, which do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man sitting at a table with a plate of food in a cup, and a bottle. front of him. The elements elements are present and match with the description. fulfilling though there might additional elements like a and a background in the background, which do not contradict the core objects.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4b65616b431848c331d5902b0717818721968570
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89f1e1aeaa3e34d61c636738e53990ecdf80be9a38132a2fa3989736274707fe
+size 148535
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..d7dae0f9a0f3c7234a34617d4fa5509ea473692d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a short, artistic piece featuring a young boy in a classroom setting. The boy, dressed in a white shirt, stands in front of a green chalkboard. His expression is serious, and he appears to be in deep thought. The background is blurred, but it's clear that the boy is in a classroom. The style of the video is simple yet effective, focusing on the boy and his surroundings. The use of color is minimal, with the green chalkboard providing a stark contrast to the boy's white shirt. The overall mood of the video is contemplative and introspective."
+    ],
+    "video_ids": [
+        "f2P-BVJyy8I_4_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Young boy in a white shirt standing in front of the green chalkboard, expression of serious deep thought.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young boy in a white shirt standing in front of a green chalkboard, and his expression appears serious and contemplative, matching the described condition. The presence of a ghostly figure whispering in the background does not contradict the core description, as it is an additional element that does not conflict with the primary subject and his expression.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a young boy in a white shirt standing in front of a green chalkboard. which his expression appears to and contemplative, which the description '. The setting of the bookly figure ining behind the background does not contradict the core description and as it is an additional element that does not interfere with the main focus and setting expression.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7c1d1ee2b4761b1ff8f02163916c2c8215454e4a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54bd7a22322bfa5f84c819a77166d5d8ce89509c7362e6db337e6f27ff03a626
+size 83189
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..53d8a5d69ca2db2538ea091b2bbbdd52e2372ae3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is sitting on the floor, surrounded by a large pile of blue and gray toy train tracks. The person appears to be sorting through the tracks, possibly organizing or assembling them. The tracks are scattered around the person, with some pieces overlapping and others lying flat. The person is holding a small blue piece of track in their hand, which they seem to be examining or picking up from the pile. The setting appears to be indoors, with a wooden floor beneath the person and a white wall in the background. The overall style of the video is casual and candid, capturing a moment of play or work with the toy train tracks."
+    ],
+    "video_ids": [
+        "YXFhzhbNaSc_5_24to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, blue and gray toy train tracks, a small blue piece of track.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person sitting on the floor surrounded by blue and gray toy train tracks, and the person is holding a small blue piece of track. These elements match the description exactly, and additional elements (like other toys or packaging) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a person interacting on the floor surrounded by blue and gray toy train tracks. which the person is interacting a small blue piece of track. The elements match the description of, fulfilling there elements likelike the toys or the) do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..10160884803dec1ee596e1a6b6b85704839cbc84
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1352-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:512917d46d3e705e9e47ba9b9ae298a2e2223e4d417cebd681b2f1f919879f86
+size 160561
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..25a9513edf733fea207d9f255e9eb36f6aae4359
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a lively scene at a movie premiere. A group of four people, two women and two men, are standing on a stage. The man in the center is holding a microphone, suggesting he is the host or a speaker. The backdrop is a vibrant blue, adorned with white geometric shapes, adding a modern touch to the event. The attendees are dressed in stylish outfits, with the women wearing dresses and the men in jackets, indicating a formal occasion. The overall atmosphere is one of excitement and anticipation, as the group seems to be waiting for the event to begin."
+    ],
+    "video_ids": [
+        "ClqkkWtzNO8_9_18to139"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Four people (two women and two men) standing on a stage. A man in the center holding a microphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows five people standing on a stage, which includes two women and three men. The man in the center is holding a microphone, which matches the description. The presence of an additional person does not contradict the core description, as the condition only requires four people (two women and two men), and the additional person is not a contradiction since the condition does not specify an exact number beyond 'four people'.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows four people standing on a stage, with slightly two women and two men. However man in the center is holding a microphone, and ful the description. The presence of an additional man does not contradict the core description as as the main only specifies four people,two women and two men) and the video person does acceptable a contradiction.\"\n the core allows not specify the exact number of fourfour'.'.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..881e1572b6c7a7acc1bf0a471cec2d4cc95e13b5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6e60e33471b2f881001b7f72c219a219425a786900ec46683c14b23f9e4a882
+size 229088
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..0cab3bd18d34b0ae9f9d74916a2b15fc89fdde02
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a computer tower with a vibrant, colorful LED light display on its front panel. The tower is placed on a desk in a room with a modern, tech-savvy aesthetic. The room features a colorful geometric pattern on the wall, and there are shelves filled with various tech-related items in the background. The video captures the essence of a high-performance gaming or workstation setup, emphasizing the visual appeal of the LED lights and the overall ambiance of the room."
+    ],
+    "video_ids": [
+        "gO-V8E9MIBg_109_0to112"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A computer tower with a vibrant, colorful LED light display on its front panel, shelves filled with various tech-related items.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a computer tower with a vibrant, colorful LED light display on its front panel, as described. The background includes shelves filled with various tech-related items, which aligns with the description. The presence of additional elements like the colorful wall lights and other decor does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a computer tower with a vibrant, colorful LED light display on its front panel, which described. The shelves includes shelves filled with various tech-related items, which alsos with the second. The presence of these elements like the shelves background and and the tech does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d972342b5a10abe9589ebbb4d2b63a6ddb718626
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c2d2e1b60c083d540141d1a41eb7369a5a08a2fca9d78cceb7d99d04af991751
+size 65811
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..642c3750ab640227d5ac9696286be72cc78360e6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a football player in a white and yellow uniform, smiling broadly and looking up towards the sky. The player is wearing a black headband with a logo on it. The background is dark and out of focus, emphasizing the player as the main subject. The style of the video is a close-up shot with a shallow depth of field, focusing on the player's face and upper body. The lighting is bright, highlighting the player's smile and the details of the uniform. The overall mood of the video is positive and uplifting, capturing a moment of joy and accomplishment."
+    ],
+    "video_ids": [
+        "LsBEeCj2iFo_2_0to119"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Football player in white and yellow uniform, black headband with a logo? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a white and yellow uniform with black accents, and a black headband featuring a logo (Pittsburgh Steelers logo). The description is accurately fulfilled without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close player wearing a white and yellow uniform, a accents, which a black headband with a logo.whichumasburgh Steelers).). The player matches largely represented by any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..db02e37e9bcd5ca443f6d3c9a1b1c4f65c2e2c6a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da133d6dd3473afb9ecd029631714bca71f5cbb809e1d2b240d0f5c3b3f5a899
+size 205821
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..60ab3b3b496f50a519783e8a0389529d47692369
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man on a boat, holding up a smartphone displaying a speedometer reading of 21 mph. The man is wearing a cap and glasses, and he appears to be in motion, as suggested by the speedometer reading. The background features a body of water, possibly a lake or river, with trees and a cloudy sky. The style of the video is casual and seems to be taken from a first-person perspective, capturing the man's experience of speeding on the water."
+    ],
+    "video_ids": [
+        "rnp9JZSMC8Q_22_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man on a boat, a smartphone displaying a speedometer reading of 21 mph.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man on a boat, holding up a smartphone that clearly displays a speedometer reading of 21 mph. The man is wearing appropriate attire for being on a boat, and the background shows water and trees, consistent with being on a boat. The core elements described are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man on a boat holding holding a a smartphone that displays displays a speedometer reading of 211. The background is wearing a attire for bo on a boat, and the background shows a and a, which with a on a boat. The smartphone elements of in present represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6c2c322c460ab3d83e5eaad0413dd0b6649145ad
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2a1eadd31da8c54ca0f27cd67d75b3a81d1fa44cd7fefdce696a9cb502637762
+size 178874
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..dc5ed79ff82bf72cc07be8d15369fa1a120470de
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features two men in a room with a window. The first man is wearing glasses and a white shirt with suspenders. The second man is wearing a purple shirt and a black belt. Both men are standing and looking to the side. The room has a window with curtains. The style of the video is a sitcom, with a focus on the two men and their expressions."
+    ],
+    "video_ids": [
+        "8acoq2i9xl4_49_23to177"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, one wearing glasses and a white shirt with suspenders, the other wearing a purple shirt and a black belt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men standing side by side. The man on the left is wearing glasses and a white shirt with suspenders, matching the description. The man on the right is wearing a purple shirt and a black belt (specifically, a black fanny pack or waist belt). The core description is accurately represented, with no significant contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men standing side by side. The man on the left is wearing glasses, a white shirt with blackers, while the description. The man on the right is wearing a purple shirt and a black belt,withally, a black beltanny pack, belt bag, The description elements is largely represented in with the significant contradictions or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e7e88785db55f18252e11d9f8d85df5558b85bfb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86129125d09764ef245d2da2c74f7ad1726996e2d681431b9e2309b8cee81f46
+size 82225
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..fd03f2c306f39df72a8903e89075a051e08887b8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a blue car with gold rims driving on a snowy road. The car is shiny and appears to be in motion. The road is covered in snow, and there are tire tracks visible in the snow. The car is the main focus of the video, and it is driving towards the camera. The background is mostly white due to the snow, but there are some trees visible. The car is the only object in motion in the video. The style of the video is realistic, and it captures the car in a dynamic and active state."
+    ],
+    "video_ids": [
+        "5PAPbm8Ez3E_7_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue car with gold rims? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a blue car with gold rims, which matches the core description. The car's color is distinctly blue, and the wheels have a visible gold finish. Additional elements like snow and bleachers do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows depicts a blue car with gold rims, which matches the description description. The car is color and blue blue, and the rims have a metallic gold rim, The elements like the and aachers in not contradict the main but}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c6f0481a25deaf8b939cce85f04b1a2ce22b9008
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11dc0831872c6b6a20148a2d89cc488aa6ce2deb71e93fcdb8139631d81c42a0
+size 344325
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..0d08fc92b50de65c741d4633d3a5140c38169bd4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a motorcycle dashboard, focusing on the speedometer. The speedometer is black with white markings and a needle indicating the current speed. The motorcycle's handlebars are visible in the foreground, with various controls and switches. The background is blurred, but it appears to be an indoor setting with a wooden floor. The style of the video is a straightforward, close-up shot with a shallow depth of field, emphasizing the speedometer and the motorcycle's controls."
+    ],
+    "video_ids": [
+        "-Fi56RPKZ9s_12_57to197"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Motorcycle dashboard, speedometer, handlebars, controls and switches.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a close-up of a motorcycle's dashboard, including a speedometer with a needle and digital display, as well as handlebars with visible controls and switches. All elements mentioned in the condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a close-up of a motorcycle dashboard dashboard, which a speedometer, a needle pointing various read, handle well as handlebars and controls controls and switches. The elements described in the ' are present and accurately depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..67073c0a0f3bfd397d5c71d53d79b9210789f2ee
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d7ff1adfaf8d461df1f18fa1caf845d6b7c778170c0096bc9a20919aa612325d
+size 143778
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..d0a559c9c0d6e0658143bbe89b8ef7699924bb24
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the growth of three sunflowers in a garden. The first frame shows the sunflowers in their early stages, with small buds beginning to form. The second frame shows the sunflowers in full bloom, with bright yellow petals and brown centers. The third frame shows the sunflowers in their final stages, with the petals beginning to wilt and the centers becoming more prominent. The sunflowers are surrounded by lush green foliage, and the garden appears to be well-maintained. The video is a time-lapse, capturing the growth of the sunflowers over time."
+    ],
+    "video_ids": [
+        "YcYVOik4HGc_1_0to103"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three sunflowers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows three sunflowers in the foreground, which matches the 'Object(s)' condition. Additional elements like bees and background foliage do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows shows three sunflowers, a sequence, with align the 'Object(s)' condition. The elements, the and the foliage are not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..16ace52be5960264699b408e569a3c3a3b4b8515
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1353-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40b4a59736c4c735de323648f5bfcc66a00506fe09cd899f94ed5231fa0be99c
+size 160359
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..668e0b5458b6e00e3e73ce5d2b1a0ffde9d78983
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a white BMW convertible car, taken from the perspective of the passenger seat. The car is parked in a serene, wooded area with a clear sky. The convertible top is down, revealing the open sky above. The car's interior is well-lit, with the sun shining through the windows. The dashboard and steering wheel are clearly visible, and the car's seats are white, matching the exterior. The car's interior is clean and well-maintained, reflecting the car's luxury and sophistication. The video captures the essence of a leisurely drive through a beautiful, natural landscape, with the convertible car as the focal point."
+    ],
+    "video_ids": [
+        "ZJ97khxv0oY_17_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: White BMW convertible car, dashboard, steering wheel, car's seats? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the interior of a white BMW convertible car, including the dashboard, steering wheel, and car's seats. All specified objects are present and accurately depicted without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a interior of a white BMW convertible car, including the dashboard, steering wheel, and car's seats. The the elements are present and match depicted, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7802b49416c14cd869ece2d691c05ee97bc4dd11
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5799d804b9d62d6f143638ac874200c2b01d20481478931f44c3b7c61daf9e7
+size 237834
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..eca8d77597b1cfacdab2460e0063f78dc300e7c6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up view of the interior of a luxury sports car, showcasing the steering wheel and dashboard. The steering wheel is black with red stitching and features the iconic Ferrari logo in the center. The dashboard is adorned with various gauges and controls, including a speedometer and a tachometer. The car's interior is sleek and modern, with a focus on performance and luxury. The video captures the essence of the Ferrari brand, highlighting its attention to detail and commitment to delivering an exceptional driving experience."
+    ],
+    "video_ids": [
+        "ZTCkq45b2BU_21_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, dashboard, gauges, controls (speedometer, tachometer).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the steering wheel with the Lamborghini logo, the dashboard, and digital gauges including a speedometer and tachometer. The controls on the dashboard and center console are also visible, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a steering wheel, the Ferrariorghini logo, the dashboard with and various gauges including a speedometer and tachometer. The presence on the steering, steering console are also visible, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..152c8be4ef350d852c05c2385372991fd689ba36
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d0a1b4a0530b0b294a7d8b4b48311ad7801b1c919a8877091a1e03a294fae833
+size 219583
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..84f3104614db3caa7d75cce72badc1fce756d6b8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car from the perspective of the passenger seat. The car is a modern SUV with a beige interior. The seats are upholstered in a light beige color, and the car features a touch screen display on the center console. The steering wheel is on the right side of the car, indicating that the car is designed for driving on the right side of the road. The car appears to be in motion, as suggested by the slight blur of the background. The video is likely taken during the day, as the interior of the car is well-lit. The style of the video is straightforward and documentary, focusing on the car's interior without any additional embellishments or distractions."
+    ],
+    "video_ids": [
+        "CHHpkZWff3g_6_0to127"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Touch screen display on the center console, steering wheel on the right side.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a touch screen display on the center console and a steering wheel positioned on the right side, which is typical for vehicles designed for right-hand driving. These elements are accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a car screen display on the center console and a steering wheel on on the right side, which align consistent for vehicles in for right-hand drive. The elements are consistent depicted in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6970f64c4d2a32510b6930ecd9b1fda0851181ae
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0fb35f2500a113e3d9f06436b3477d2f4c27a5d34485c4e6cca20994fb9e1065
+size 244465
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..fe4bf34bb551d90724bb0933d848e80a7d5a41a9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a picturesque village nestled in a valley surrounded by mountains. The village is composed of quaint, colorful houses with red roofs, and a dirt road winds its way through the center. The houses are closely packed together, creating a sense of community and togetherness. The village is surrounded by lush greenery, and the mountains in the background add a sense of tranquility and seclusion. The sky is a clear blue, and the sun is shining brightly, casting a warm glow on the village. The overall style of the video is serene and idyllic, capturing the beauty of rural life."
+    ],
+    "video_ids": [
+        "wOxB9PzJNLE_55_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Quaint, colorful houses (red roofs), a dirt road, lush greenery, mountains, and a clear blue sky with a bright sun.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by showcasing quaint, colorful houses with red roofs, a dirt road winding through the scene, lush greenery with autumnal hues, prominent mountains in the background, and a clear blue sky with scattered clouds and bright sunlight. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition by depicting a, colorful houses with red roofs, a dirt road, through the village, lush greenery, treesal trees, mountains mountains in the background, and a clear blue sky with a clouds and a sunlight. The elements elements of in present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..56e7b75d021866dc26c3dee96be72667a1289807
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da3dbf0337c9942a8d0327eca6c0bb72b16b13922af8b1ad36411931f5f4146f
+size 201480
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..2baed93453561247bcc4e3e0227e2f1157559040
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a white Mustang sports car with black and red stripes on its hood, parked in a parking lot. The car is positioned in the center of the frame, with the camera angle slightly elevated, giving a clear view of the car's front and side profile. The car's sleek design and shiny exterior reflect the surrounding environment. The parking lot appears to be empty, with the car being the only vehicle visible. In the background, there's a building and a palm tree, suggesting a warm climate. The lighting in the video is soft, indicating that it might be either early morning or late afternoon. The overall style of the video is clean and minimalist, focusing solely on the car and its surroundings without any additional elements or distractions."
+    ],
+    "video_ids": [
+        "UjjTjf8Kyzk_31_0to117"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white Mustang sports car with black and red stripes on its hood? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white Mustang sports car with distinct black and red stripes running down the hood, matching the description. The camera slowly zooms in on the front of the car, highlighting these stripes and the car's design. No elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white Mustang sports car with black black and red stripes on along the hood, which the description provided The car angle zooms in, the car of the car, providing the features. the overall's sleek. There additional in the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f9bf5d2f4108e776eebfe2894ea7d8fd0e175d81
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e7722ef2d513ab5af7eef74bfcf60a1e68b6b3566377dcef1405c4d5745cec7
+size 163773
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..7e79c4a0ca125ef0347d63036c2f8f50496ba0d0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene day at a marina, where a sleek black speedboat is docked. The boat, with its polished exterior and large windows, is the focal point of the scene. It's moored next to a white dock, which extends into the calm blue water. The marina is bustling with activity, with several other boats visible in the background. The sky above is a clear blue, suggesting a sunny day perfect for boating. The overall style of the video is a simple yet beautiful depiction of a day at the marina, with the speedboat as the star of the show."
+    ],
+    "video_ids": [
+        "QmJY9PtPp5M_45_0to162"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A sleek black speedboat, a white dock, several other boats.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a sleek black speedboat docked at a white pier, with several other boats visible in the background. The scene matches the described elements without contradiction, even though additional elements like a blue mat and a white inflatable boat are present, which do not conflict with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features a sleek black speedboat,ed in a white dock. which several other boats visible in the background. The description is the description elements without any.\"\n fulfilling though the boats like the mar sky on a white boat boat are present, they do not conflict with the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0756b28b271c8446f652acbc34fa266b9046dbfe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd987b016e23909c09a8aa696510a8014af4fc86d1f023ddcc8b2aaf34c2348a
+size 190372
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..59681f527e5b327007edf7cf0f5f7489804f6f45
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene by a pond where a fluffy brown cat is intently watching a group of koi fish swimming gracefully in the water. The cat, positioned on a stone ledge, leans forward, its attention fully captivated by the vibrant colors and movements of the fish. The koi display a variety of hues, including white, orange, and yellow, with some featuring intricate patterns on their scales. The water is clear enough to see the fish clearly, and small ripples form around them as they swim. The background reveals a tranquil setting with a mix of greenery and a calm surface of the pond. The camera remains steady throughout the sequence, focusing on the interaction between the cat and the fish, highlighting the peaceful coexistence of the two animals in this natural environment."
+    ],
+    "video_ids": [
+        "8d723d214357a4b0889207b2bd05192d832ed55fcee3e61d6207cfaf71f67f83"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A fluffy brown cat and a group of koi fish.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a fluffy brown cat sitting by the edge of a pond, observing a group of koi fish swimming in the water. The cat's fur, color, and posture are consistent with the description, and the koi fish are visible and active in the water. No elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a fluffy brown cat and on a edge of a pond, observing a group of koi fish swimming in the water. The presence and fur is the, and the, consistent with the description of and the presenceoi fish are visible in swimming in the pond, There additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..47bc7b2a5ad6007ecd7690f1c51fe620afb29b5d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c02625554ca3eb7eb5027cfba403942dd040a78c875c3ad604f545d85516838b
+size 215206
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..1e98606aa3d5a1d7693d9723aebb1e748a65dc82
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man in a red superhero suit with a yellow emblem on his chest is seen standing on a circular platform with blue lights. He has his arms outstretched and is surrounded by a halo of blue light. The platform is located in a room with a gray floor and walls. In the foreground, there is a woman with blonde hair and a black and white costume. She is standing with her hands on her hips and is looking at the man on the platform. The overall style of the video is animated and has a futuristic feel to it."
+    ],
+    "video_ids": [
+        "JWyUltNDgrw_21_25to188"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a red superhero suit with a yellow emblem, a woman with blonde hair and a black and white costume? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a red superhero suit with a yellow emblem (Mr. Incredible) and a woman with blonde hair wearing a black and white costume (Elastigirl). Both characters are accurately represented as described, with no conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a man in a red superhero suit with a yellow emblem,consistent. Fantastic) and a woman with blonde hair wearing a black and white costume.Vlastigirl). The characters are present depicted as described in and the additional elements present}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..906aee67880d7060ee04173cfc86bd91fb53ae78
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1354-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a8a84c9d27d9b4856e4e41e575bc15984bda9b2d64a32eba5cb54fe0a325deb8
+size 136278
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..13325e6a7db91dfa8c4abfde208d34724e1e30ff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the raw beauty of a mountainous coastline. The first frame shows a large wave crashing against the rocky shore, with the mountainous backdrop partially obscured by the spray of the water. The second frame shows the wave receding, revealing the rugged cliffs and the mountainous backdrop in more detail. The third frame shows the wave crashing again, with the mountainous backdrop now fully visible. The style of the video is dynamic and dramatic, capturing the power and beauty of nature."
+    ],
+    "video_ids": [
+        "m23npIAgEDI_25_0to135"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Waves, rocky shore, rugged cliffs, mountainous backdrop? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by showcasing waves crashing onto a rocky shore, with rugged cliffs and a mountainous backdrop visible in the background. The transition from a glacier-fed lake to a turbulent ocean scene with waves and rocky shorelines aligns with the described elements, even though the initial scene is not oceanic. The core description is met without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting waves crashing onto a rocky shore, rugged rugged cliffs and a mountainous backdrop. in the background. The sequence between calm calm to river to a rocky ocean scene with waves crashing foam cliffslines furthers with the description elements, indicating though the initial scene is not explicitlyic. The video elements of met with any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6712eca4f1ba080abce287d3c9ee77011149dfff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9324cc2d8a688e1a4f047f57b6daa8bc77b3db5295d40cf5533ec2071fc5e9f8
+size 246699
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..200d8de126c0fa155ab4c766158e9c4dd4c7501e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a black sports car parked on a paved surface, possibly a parking lot or a street. The car has a sleek design with a low and wide stance, featuring large black wheels with orange accents. The car's body is shiny, reflecting the surrounding environment. In the background, there are other vehicles parked, including a white van and a red car. The setting appears to be an urban or suburban area with a building and a fence visible in the distance. The style of the video is straightforward, with no visible text or additional graphics. The focus is on the car, and the camera angle is slightly elevated, providing a clear view of the vehicle's design and color. The lighting suggests it might be daytime, and the overall atmosphere is calm and quiet."
+    ],
+    "video_ids": [
+        "dPr7LOjIKVU_31_0to146"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Black sports car, white van, red car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a black sports car, which matches the core description. In the background, a white van and a red car are also visible, fulfilling the additional elements mentioned in the condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a black sports car, which matches the description description. Additionally the background, there white van is a red car are also visible, fulfilling the ' elements of in the condition. conflicting.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8fce0e6a210f0655b4c11549c89c389e4acb6929
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07e6d79b462ccc572f580b62028c51f72ed1feee91f266a461a33c35c329a775
+size 244613
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..2add038e15bcacb2a387712e8d84fdda44bc1cbe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a soccer match. The central figure, a player in a black jersey, is in motion, running towards the camera. His arms are outstretched, suggesting he is either celebrating a goal or preparing to make a move. The player's jersey is clearly visible, with the word \"Fly Emirates\" prominently displayed.  In the background, two other players can be seen. One is wearing a white jersey with green stripes, while the other is in a white jersey with a green logo. They appear to be in pursuit of the central player, adding to the sense of action and competition in the scene.  The setting is a soccer stadium, with the stands visible in the distance. The lighting suggests it's either late afternoon or early evening, casting a warm glow over the scene. The focus is on the players and their actions, with the background slightly blurred, emphasizing the intensity of the moment.  Overall, the video is a snapshot of a high-energy moment in a soccer match, capturing the speed, skill, and competition of the sport."
+    ],
+    "video_ids": [
+        "Jnjsa1V7_iY_10_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a black jersey, a player in a white jersey with green stripes, and a player in a white jersey with a green logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a black jersey (Neymar) surrounded by two players in white jerseys with green stripes (Celtic FC players). One of the white jerseys also has a green logo visible on the back. The description accurately matches the visual content of the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two player in a black jersey andlikelyottsmar Jr and by two players in white jerseys. green stripes andRealristic players players). The of the Celtic jersey also has a green logo,. the side. The presence matches matches the core elements of the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0a7711a99d377ac6f8335e9cc26f7389bca79656
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d2e1b454a1ab2954c0a476dfe33d2d6bd94db135421676fd433b2e29121ecff4
+size 154906
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..1e83ebbdc2c6b8cfbf0fc5c53320c25b04a36ec8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with long blonde hair, wearing a white baseball cap and a black jacket. She is seated in a room with a window covered by gray blinds. The woman appears to be looking off to the side, her expression neutral. The room has a simple, uncluttered appearance, with the blinds being the only notable feature in the background. The lighting in the room is soft and even, suggesting an indoor setting. The woman's attire and the casual setting suggest a relaxed, informal atmosphere."
+    ],
+    "video_ids": [
+        "CJ8MsaoY7Ic_0_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with long blonde hair, wearing a white baseball cap and a black jacket. She is seated and looking off to the side with a neutral expression.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with long blonde hair, wearing a white baseball cap and a black jacket, seated and looking off to the side with a neutral expression. These elements match the description provided in the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with long blonde hair, wearing a white baseball cap and a black jacket. seated and looking off to the side with a neutral expression. The elements match the description provided, the question.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4c4ade31a825303c8833757a3d62e32aedc88506
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2607a3dcf6ea8fc6bd0f30fd67d5d404456a3fc760244b53c1d6b138bf6d7e04
+size 99537
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..136064ec15817a327cfcaf5795878f614d63b443
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a man in a suit and tie discussing the topic of \"Open Carry\". The man is standing in front of a blue background with a news ticker at the bottom. In the corner of the screen, there is a small inset image showing a person's backside with a handgun in a holster. The style of the video is a typical news report with a professional and serious tone. The man is speaking directly to the camera, and the inset image is used to visually represent the topic being discussed. The overall impression is that of a well-researched and informative news segment."
+    ],
+    "video_ids": [
+        "kKBpc2xUqUo_4_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, a small inset image of a person's backside with a handgun in a holster.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a suit and tie speaking, and there is a small inset image of a person's backside with a handgun in a holster, which matches the described condition. The additional elements (e.g., news studio background, on-screen text) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a man in a suit and tie,, which there is a small inset image of a handgun's backside with a handgun in a holster. which matches the description '. The presence elements inthe.g., the ticker background, text-screen text) do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..54fc1c796f78904150cd2aeb83cc1da5606682ac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9aa4bdf61ef4bfc5a78d9cd5fd72ec51da63a981f2c0baf812358d5dad70fb2
+size 92465
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..0398173706b4fb3280191d573e1e6e987d100fc4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the breathtaking beauty of a mountainous landscape, where a waterfall cascades down a rocky cliff into a serene river. The river, filled with rocks, flows gently, reflecting the clear blue sky above. The waterfall, a stunning spectacle, adds a sense of motion to the otherwise tranquil scene. The mountains, majestic and towering, provide a stunning backdrop to the waterfall and river. The video is shot in a way that emphasizes the natural beauty of the landscape, with a focus on the waterfall and river, while the mountains serve as a constant presence in the background. The overall style of the video is one of tranquility and natural beauty, capturing the essence of the landscape in a way that is both visually stunning and emotionally evocative."
+    ],
+    "video_ids": [
+        "0tCPQ17cSpI_24_16to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Waterfall, rocky cliff, river, clear blue sky, mountains? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a waterfall cascading over rocks, a rocky cliff forming the backdrop, a river flowing through the scene, a clear blue sky above, and prominent mountains in the background. All specified elements are present and accurately depicted without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a waterfall cascading down a, a rocky cliff, the backdrop, a river flowing through the scene, a clear blue sky,, and mountains mountains in the distance. The these elements are present and contribute depicted in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0e4ac7ca8e9d492b8b9674870ce1bee39b884a80
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:942f563aa51089e712494be99df3d026d0f0b6aeef4042f2f5f3b06c30c4513e
+size 162238
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..ff9816a6a8e4187c66cfe0b4ea58305b37b191f5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a car's interior, focusing on the engine start button. The button is illuminated with a red light, indicating that the car is in the process of starting. The car's interior is sleek and modern, with a black dashboard and silver accents. The driver's hand is visible, pressing the button to start the car. The video captures the anticipation and excitement of starting a journey in a new car. The style of the video is realistic and straightforward, focusing on the car's interior and the action of starting the engine."
+    ],
+    "video_ids": [
+        "_zMeMk7OJE0_17_0to146"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Engine start button, driver's hand? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows an engine start button labeled 'ENGINE START STOP' and a driver's hand pressing it. The button lights up red after being pressed, and the car's climate control display becomes active, indicating the engine has started. These elements align with the described 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a engine start button, withSTART START'' and a driver's hand interacting it. The focus and up,, being pressed, which the hand's interior control system is visible, indicating the engine has started. The elements fulfill with the description 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..72f65b72980734dc5441664d086bb90397d8355d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:914fc4fe1d608282c24067af38ea6feb7c73916a69c79e5305e2b3665c8e5991
+size 191254
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..08a224f7c684a93d38195f51424c2c8d8242af66
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of an elephant standing in a grassy field enveloped by a thick layer of fog. The elephant, with its large ears flapping gently and tusks prominently visible, is positioned centrally in the frame. Its body is slightly angled towards the camera, giving a clear view of its massive form. The fog creates a soft, ethereal atmosphere, obscuring the background and adding a sense of mystery to the setting. The elephant appears calm and stationary, occasionally moving its trunk or ears slightly. The grass beneath it is lush and green, contrasting subtly with the muted tones of the fog. The overall mood of the video is tranquil and contemplative, emphasizing the majesty of the elephant in its natural habitat. There are no other animals or significant changes in the environment throughout the sequence, maintaining a consistent and peaceful visual narrative."
+    ],
+    "video_ids": [
+        "126791068fda5139f6b7ea662b4009787ea49aff79e6a3e3a7576564fc657b00"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An elephant? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows an elephant standing in a misty field, which matches the 'Object(s)' condition of an elephant. The elephant's features, such as its large ears, trunk, and tusks, are visible despite the fog, and there are no elements contradicting the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts an elephant, in a grassy environment. which directly the descriptionObject(s)' condition of the elephant. The presence is features, such as its large ears, tus, and tusks, are prominently and the foggy confirming the are no additional ining the presence of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2b73b7dc2fbb6ba0c72e8abf7a7fbe1837500b87
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1355-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7507cd822393f23605bb6c0d46e77689d273d075ad737462d4c735719b62d571
+size 161574
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..907b19db09b385dbd63f90d8258b29d1d905b44b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a small monkey, possibly a Kinkajou, in its natural habitat. The monkey is seen climbing a tree, gripping onto a branch with its small hands. The monkey's fur is a mix of brown and white, and it has a curious expression on its face as it looks directly at the camera. The background is filled with lush green leaves, suggesting that the scene takes place in a dense forest or jungle. The monkey's movements are slow and deliberate, indicating that it is carefully navigating the tree. The overall style of the video is a close-up, slow-motion shot that allows the viewer to appreciate the monkey's features and the surrounding environment."
+    ],
+    "video_ids": [
+        "TLs_kV4Zf3w_23_112to248"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small monkey (possibly a Kinkajou)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a small, brown-furred animal hanging from a tree branch, which matches the description of a Kinkajou. The animal's features, including its round face, large eyes, and furry body, are consistent with known characteristics of a Kinkajou. The text overlay also explicitly identifies the animal as 'Kinkajou', confirming the object's identity.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a small pr furry andurred animal with from a tree branch, which align the description of a smallinkajou. The animal's size, such its fur face, small eyes, and small body, align consistent with the characteristics of a Kinkajou. The settingural in suggests identifies the animal as aKinkajou', further the match condition identity.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e17f1df164315fbfd4bbe1cee6c842e85406e713
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:517ad10444433010d1f39cfcbbeea619fbf9d4de01f62108a8e0cb6e52a250d7
+size 177553
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..5c7eb912b6fd475739c56c4ab39d60be7c17adde
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a kitchen, holding a plate of food. She is standing in front of a counter, which has a bowl and a vase on it. The kitchen is equipped with modern appliances, including a microwave and an oven. The woman is dressed in a black jacket and a floral dress. The video captures the woman's journey from preparing the food to serving it on the plate. The overall style of the video is realistic and it captures the everyday life of a person in a kitchen."
+    ],
+    "video_ids": [
+        "ySCSeITKRzk_11_0to142"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a plate of food, a counter, a bowl, a vase, a black jacket, and a floral dress.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a kitchen holding a plate of food. A counter is visible with a bowl and a vase on it. The woman is wearing a black jacket over a floral dress. All specified objects are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a woman in a kitchen setting a plate of food. The counter is visible, a bowl and a vase on it. The woman is wearing a black jacket over a floral dress. The the elements are present and the with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b1f7f61bcc359fab243ab8485ada59cdb914a027
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a2a88d6ce54d396e71c086ed3bace8e83db770e8fd8753fbc7801933dcaee1b8
+size 165967
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..6cc969aff0932dc0f349db3fafb71fa02b168c0e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a grill filled with various types of meat and vegetables being cooked over an open flame. The scene is set against a dark background, which makes the vibrant colors of the food stand out prominently. The grill is filled with skewers of meat, including what appears to be chicken wings and possibly some beef or pork, along with some green vegetables, likely bell peppers or zucchini. The meat is sizzling and browning, indicating it's being cooked at a high temperature. Flames lick at the edges of the meat, adding to the visual appeal and suggesting the heat is intense. The camera remains stationary throughout the sequence, focusing on the grill and the cooking process. The lighting highlights the textures and colors of the food, making the scene look appetizing and lively. There are no visible characters or significant changes in the environment; the focus remains solely on the grilling process."
+    ],
+    "video_ids": [
+        "92d905af0171de591effbd32ceafe2658a122fbb0935e612b5a0e2a662619388"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Grill, skewers of meat (chicken wings, beef/pork), green vegetables (bell peppers/zucchini).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a grill with skewers of meat (appearing to be beef or pork) and green vegetables (likely bell peppers) being cooked over flames. The visual elements align with the described objects, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows skew grill with skewers of meat,whicharing to be chicken or pork) and green vegetables (bell bell peppers and being cooked. an. The presence elements match with the description objects: fulfilling there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7c19a18255082ee1575506060945f6d95e9a32cb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57eaae609135022c21cc4c378e247b2beaef7991b5afd0ed9cf6e51da43b4295
+size 117907
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..57430720662c913b779106f7fde6e62f2efd432f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a gray Jeep Cherokee SUV on display at an auto show. The car is positioned on a raised platform, allowing viewers to appreciate its design and features. The vehicle is equipped with a black roof rack, adding to its rugged appeal. The Jeep Cherokee is parked on a gray platform, which contrasts with the car's color and highlights its presence. The background is dark, ensuring that the focus remains on the car. The overall style of the video is sleek and professional, typical of auto show presentations. The video does not contain any text or additional elements, keeping the viewer's attention solely on the Jeep Cherokee."
+    ],
+    "video_ids": [
+        "teJXEp68_pM_14_0to119"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Gray Jeep Cherokee SUV, black roof rack? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a gray Jeep Cherokee SUV with a black roof rack. The vehicle is prominently displayed, and the roof rack is visible on top of the SUV, matching the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a gray Jeep SUV SUV with a black roof rack. The vehicle is positioned displayed on and the black rack is visible, the of the SUV, matching the description provided.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bf22f2ae7a611cfb5aeb86a75e0266f9c529283d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45dc5a959b0f604c963a70284f1c8fe6d13857b2260e86d6c6d9c2306e5102cc
+size 143515
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..3d0dd478d9494da6f3805610940162e0caafc380
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a white sailboat floating on a clear blue ocean. The boat, with its red flag fluttering in the wind, is the main focus of the video. The sailboat is seen from a distance, giving a sense of its size and the vastness of the ocean. The boat is moving towards the right side of the frame, suggesting a journey or adventure. The background is filled with the tranquil beauty of the ocean, with a few other boats visible in the distance. The overall style of the video is calm and peaceful, capturing the essence of a day out on the water."
+    ],
+    "video_ids": [
+        "9ngJMDB7cQw_1_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white sailboat with a red flag.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a white sailboat with a red flag on its stern, which matches the core description. Additional elements such as solar panels and another sailboat in the background do not contradict the primary object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a white sailboat with a red flag, its mast, which align the description description. The elements such as the panels on a sailboat in the background do not contradict the main object description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..65ad166fe89246066cc80de7ee14d6e2adad5aac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:875d3060e4f363d09215f3f0d20364496ce7a2acf140d9556db898ddab0fcc63
+size 212729
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..3f603341a94ceb2f63b8cbc5806c34485c2168eb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video opens with a close-up shot of a small, light-colored animal, likely a pika or similar rodent, moving swiftly across a rocky terrain. The animal's fur blends well with the gray and white rocks, making it initially difficult to spot. As the camera pans out, we see the animal running along a path lined with large boulders and patches of green grass. The background reveals a rugged mountainous landscape under a clear blue sky, suggesting a high-altitude environment. The camera follows the animal as it navigates through the uneven ground, occasionally pausing to sniff the air or look around. The scene captures the natural agility and alertness of the creature in its habitat, emphasizing the stark beauty of the alpine setting."
+    ],
+    "video_ids": [
+        "b10a8db608c66e1196bea95aaa755740d60d381173bf3297586b022c03150d91"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small, light-colored animal (likely a pika or similar rodent).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video consistently shows a small, light-colored animal moving through a rocky, grassy environment, which matches the description of a pika or similar rodent. The animal's size, coloration, and behavior are consistent with the expected characteristics of such an animal.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a small, light-colored animal that through a rocky terrain grassy environment. which align the description of a pika or similar rodent. The animal's size, color,, and the align consistent with the characteristics characteristics of a an animal.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d6a60f2355e2f1451e2cf34ae29b45f581ce41f8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2ec91141dd00207468c376e0ea5c831a85a34487ff0fbf9c1e78b71cf51ceefb
+size 314403
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..611e04904bdef5f88c0585cc9f757a7872d5e635
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car, focusing on the driver's side. The car has a beige interior with a black steering wheel and a black dashboard. The steering wheel has a logo on it, indicating the make of the car. The dashboard has a radio and air conditioning vents. The car has a manual transmission, as evidenced by the gear shift lever. The car's interior is clean and well-maintained. The video is a straightforward, unembellished depiction of the car's interior, with no additional context or action. The style of the video is straightforward and unembellished, focusing solely on the car's interior."
+    ],
+    "video_ids": [
+        "XHZmNU-De3A_52_0to143"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, dashboard, gear shift lever? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the steering wheel, dashboard, and gear shift lever as requested. These elements are prominently visible and accurately represented without any contradictions. The presence of other interior components like seats and floor mats does not conflict with the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a steering wheel, dashboard, and gear shift lever, described. The elements are prominently featured and match depicted, any contradictions or The presence of additional elements car like the and air mats does not detr with the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a7de25b771d9c02e86c65831796479553c7ceb0c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:32956897400938ef1e16921be63250ee8f9bb030b189998458529e3921e81efc
+size 212517
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..5ea110530b9e241a6adbb3c7b1d1fbf3535f5905
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a car's wheel in motion, showcasing its design and the environment it's in. The wheel is black with silver accents and features a unique star-like pattern. The car is driving on a dirt road, with trees and foliage in the background, suggesting a rural or natural setting. The video is shot in a way that emphasizes the wheel's design and the car's movement, with a focus on the details of the wheel and the texture of the road. The style of the video is dynamic and engaging, with a sense of motion and speed conveyed through the car's movement and the camera's perspective."
+    ],
+    "video_ids": [
+        "ugNA5w0htMA_28_0to101"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black car wheel with silver accents and a star-like pattern.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a car wheel that is predominantly black with shiny silver accents forming a star-like pattern around the center. The wheel's design matches the description, and although there are additional elements like the red brake caliper and the surrounding environment, they do not contradict the core description of the wheel itself.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a black wheel that is predominantly black with silver silver accents and a star-like pattern. the hub. The wheel is design and the description provided and the the are no elements like the car car caliper and the car car, they do not contradict the core description of the wheel itself.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c62a8ad2e8b507abf4c677e6e7696ff7cbe2deca
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1356-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe86ea504be7bd9bc7347e56e193d488a1466675739c5b85fd27077b3e771678
+size 275706
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..0002ac58ba35a07d5acfb0c8244e14ce784cbe2a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a car with a unique and colorful design, featuring a pattern of cartoonish faces in various colors and expressions. The car is parked in a garage or indoor space, and the design covers the entire exterior of the vehicle. The faces are whimsical and playful, with some showing happy expressions, while others appear more mischievous or neutral. The style of the video is a straightforward, real-life capture of the car's design, with no additional elements or effects. The focus is solely on the car and its vibrant, eye-catching design."
+    ],
+    "video_ids": [
+        "x9KPot6gh8w_6_0to104"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Car with a unique and colorful design featuring cartoonish faces? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a car with a vibrant, unique design featuring numerous cartoonish faces in red, blue, yellow, and white colors, which matches the description. The faces have expressive eyes and mouths, and some include dollar signs, adding to the whimsical, colorful aesthetic. Additional elements like another car in the background do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a car with a unique and colorful design that cartoon cartoonish faces in various, green, green, and other colors. which align the description of The car are exaggerated features and mouths, adding the have additional signs, adding to the playfulical and colorful theme. The elements like the car in the background and not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..64ac655ca56c5645858a7b4c678279956ac314ac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd3bf706908fdc82ef1246adabf98090a3694ede51422be189dedd385f401a82
+size 160200
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..b1016a50dd651abcff727b5ac3a411e69c42db07
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a cooking process taking place on a stovetop. A large, dark-colored wok dominates the frame, resting on a burner. Inside the wok, a metal bowl is placed on a wire rack, holding a dish that appears to be a type of stir-fry or braised meat. The dish includes chunks of meat, possibly chicken or pork, coated in a rich, reddish-brown sauce. Red chili slices are scattered across the surface, adding a pop of color and suggesting a spicy flavor profile. A hand enters the frame from the right side, reaching into the bowl and adjusting the contents slightly, ensuring everything is evenly distributed. The background reveals a kitchen setting with various items such as jars and lids, indicating a home-cooking environment. The lighting is warm and focused on the wok, highlighting the textures and colors of the food. There is no significant camera movement; the shot remains steady, focusing on the detailed preparation of the"
+    ],
+    "video_ids": [
+        "d610483f0d3b837bd6323511a0dc110ced4bcfeaeaf149389987f395b84668dc"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large, dark-colored wok, a metal bowl on a wire rack inside the wok, chunks of meat, a rich, reddish-brown sauce, red chili slices, and a hand.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a large, dark-colored wok with a metal bowl placed on a wire rack inside it. The bowl contains chunks of meat, a rich, reddish-brown sauce, and red chili slices. A hand is also visible interacting with the contents of the bowl, confirming all specified elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a large, dark-colored wok on a metal bowl placed on a wire rack inside it. The bowl contains chunks of meat in which rich, reddish-brown sauce, and red chili slices. Additionally hand is also visible, with the food of the bowl, adding the the elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6725ba188e049b300f8d879d35c4d4176de0cdfd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c83203a61f47a88fa8b081f067eb933e06638ba6ea3a2ac778599cc9d17df1cf
+size 129364
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..7abcd7c8d395685835f81bad03d6660638e50569
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a bald man with a goatee, wearing a black polo shirt with a logo on the left chest, looking down and to the left. He appears to be in a store with framed sports memorabilia on the wall behind him. The man's expression is serious, and he seems to be focused on something out of frame. The lighting in the store is bright, and the man's reflection is visible on the glass of the display case. The style of the video is candid and unposed, capturing a moment in the man's day."
+    ],
+    "video_ids": [
+        "CD_ToYKd52A_7_0to168"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bald man with a goatee, wearing a black polo shirt with a logo on the left chest, glass display case, framed sports memorabilia on the wall.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man with a goatee wearing a black polo shirt with a logo on the left chest, which matches the description. In the background, there are framed sports memorabilia on the wall, including a display case with coins and a framed photo of football players, which aligns with the described elements. There are no significant contradictions in the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man with a goatee wearing a black polo shirt with a logo on the left chest. standing matches the description. The the background, there is framed sports memorabilia on the wall, and what bicycle case, items, other helmet picture, a players, which alsos with the description glass. The are no contradictions contradictions or the video elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..59e4311a98c0314fb0e4b299b60e949fd33006f2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:300d5259d7c5b894e33e0fc1c6f4324d59c88ceeb841b27914d757a0a38c00a7
+size 98270
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..8008aee34d6444d9392aa7c0fb62e4bea1feb065
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a vintage blue muscle car parked in front of a building with red and white walls. The car is a classic model, featuring a prominent hood scoop and a set of shiny chrome wheels. The car is parked on a paved surface, and the building behind it has a corrugated metal exterior. The car is the main focus of the video, and it is shown in three different angles, highlighting its design and details. The style of the video is straightforward and documentary-like, with no additional elements or effects. The car's position and the building's colors create a contrast that emphasizes the car's presence."
+    ],
+    "video_ids": [
+        "RKhQYKCxOwE_2_0to169"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A vintage blue muscle car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a vintage blue muscle car, specifically a Dodge Charger, with classic styling and features such as chrome trim, distinctive wheels, and a sleek body. The car is clearly the focal point and matches the description of a vintage blue muscle car.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features two vintage blue muscle car, which a classic Charger, which a features and a typical as a wheels and a headlights, and a sleek body shape The car is positioned the main point of matches the description of a vintage blue muscle car.}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2329859b893d5a6c2f1162c5156696b9c0180272
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c9845a88db83928bedc8e7c382ab2621de5eaefd717510f2d78ff5e13dd0f919
+size 267122
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..b227d0f63c81ff2b24684d8e62436db1998491b8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a black shirt with a Google Outreach badge, standing in front of a red sign with a white circle on it. He is gesturing with his right hand, pointing towards the sign. The man appears to be speaking or explaining something to the audience. The audience consists of several people, some of whom are wearing lanyards, suggesting that this might be a conference or a workshop. The setting appears to be indoors, with a plain wall in the background. The style of the video is a straightforward, unedited recording of a presentation or lecture."
+    ],
+    "video_ids": [
+        "Cn6goQPGpu8_9_0to144"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a black shirt with a Google Outreach badge, red sign with a white circle, audience consisting of several people.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black shirt with a visible 'Google Earth Outreach' badge, standing in front of a red sign with a white circle. There are also audience members visible in the foreground, matching the described elements. The scene aligns well with the core description without significant contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a black shirt with a Google GoogleGoogle Outreach Outreach' badge, standing in front of a red sign with a white circle. There is several several members seated, the foreground, wearing the description elements. The presence appearss with with the provided description provided any contradictions or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..49822a2145ad19e5b6d583d627b8044853d95771
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7fbb8451163f25c79a6a74dcc04bd970f3da578dae0880bda92aade480ff4f63
+size 126026
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..592b8dc7c891a6c1439f93c3f29768585a1f0ee4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up of a green frog in a white setting. The frog is seen in three different positions, each time with its tongue out, licking its nose. The frog's tongue is pink and long, and it appears to be in a playful or curious mood. The white background provides a stark contrast to the frog's vibrant green skin, making the frog the focal point of the video. The video is shot in a way that allows the viewer to appreciate the frog's unique features and behavior."
+    ],
+    "video_ids": [
+        "laJxNKW8x4w_5_1094to1246"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A green frog? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a green frog as the main subject, with its distinct green skin, large eyes, and extended tongue, which aligns with the description. The frog is the central focus and no elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a green frog with the main subject. with detailed distinctive green color, large orange, and a pink. which ares with the description of The presence is the central focus of there other contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9199d934bff228e1dc4e9a426c46bd1d9c78e168
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1c6bafe7862f84ded41e7745d741da7fcf6e4e77eb21c83dff86a76dcfb3d0a5
+size 107774
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..609be8e3dc4e6605428a344a2833e2c223d66fef
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a modern cityscape with a prominent skyscraper that stands out due to its unique, angular design. The skyscraper is situated near a body of water, reflecting its glass facade and creating a mirror image. The city is bustling with activity, as evidenced by the presence of other high-rise buildings, a bridge, and a bus. The sky is clear, suggesting a bright and sunny day. The overall style of the video is a blend of architectural and urban photography, capturing the essence of a thriving, modern city."
+    ],
+    "video_ids": [
+        "bKe2mx3pCew_15_149to290"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Skyscraper, body of water, other high-rise buildings, bridge, bus? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a skyscraper (the twisting tower), a body of water (river with reflections), other high-rise buildings (surrounding modern towers), and a bridge (visible in the foreground). While no bus is clearly visible, the presence of other elements does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a skyscraper,the tall tower in a body of water (the or a), other high-rise buildings ininrounding structures structures), a a bridge (span in the background). Additionally there bus is explicitly visible, the presence of a elements does not contradict the core description,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7007abe14470039da98c40e594e00caaa4eca1ae
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a56c4c9914854910eeab174fcaf03b0d6568146863ef9692c3bdab565f3b500d
+size 127091
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..d392e8b793d8ebbb45f71bf0baf1c1bad80056fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man and a woman walking together in a park-like setting. The man is dressed in a gray suit and tie, while the woman is wearing a beige coat with a bow at the waist. They are holding hands as they walk, and the woman is carrying a brown purse. The background features a brick wall and trees, suggesting an urban park or garden. The style of the video is casual and intimate, capturing a moment of companionship between the two individuals."
+    ],
+    "video_ids": [
+        "IO8Mju1COl4_7_23to226"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman walking together, which fulfills the 'Object(s)' condition. The presence of other individuals in the background does not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man and a woman walking together. which fulfills the 'Object(s)' condition. The presence of a elements in the background does not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4059b074a8c4af25abdb6a4ae06dffcea361467b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1357-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3f5822dbedd73fb7fb8f07658f19c31831bd1cc892ad045552cdd9db53c5fb4
+size 191637
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..c96756720397a2ce52ae3210b289e497ec104857
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young man with a beard and mustache, wearing a red baseball cap and a black and white checkered shirt. He is standing against a blue background with a diagonal striped pattern. The man appears to be speaking or singing, as he is holding a microphone close to his mouth. His expression is serious and focused, suggesting that he is engaged in a performance or presentation. The overall style of the video is simple and straightforward, with a focus on the man and his actions."
+    ],
+    "video_ids": [
+        "4YDrUlmIuDE_0_0to126"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: - A young man with a beard and mustache? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The young man in the video has visible facial hair, including a beard and mustache, which matches the description. His appearance is consistent with the specified criteria.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video man in the video is a facial hair, which a beard and mustache, which align the description. He attire, consistent with the ' '.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7ece5ffca7a5f807083653f18a4c88abdd640d52
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1def755ed943f506ece6fd3fc00761bafc74d72ae126f50ca00e110dffc6035e
+size 246411
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..90eb4662ab3d67445a10c370008ff29355c71326
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures an underwater scene featuring a large school of yellow fish with black stripes swimming in unison. The fish are densely packed together, creating a vibrant and dynamic visual effect as they move through the water. The background reveals a coral reef with various shades of blue and green, indicating a healthy marine environment. The camera remains relatively steady, focusing on the school of fish as they navigate through the water. The lighting is natural, filtering through the water, which enhances the colors of the fish and the surrounding coral. As the video progresses, the school of fish continues to swim in a coordinated manner, showcasing their collective behavior and the beauty of marine life. The overall atmosphere is serene yet lively, highlighting the intricate ecosystem of the underwater world."
+    ],
+    "video_ids": [
+        "40c6804112aace207eaa697a9d958be381cd27946ac108978b0f06de71044504"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large school of yellow fish with black stripes.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large school of fish that are predominantly yellow with distinct black stripes, matching the description. The fish are swimming together in a coordinated manner, forming a large group as described. While there are some other fish visible in the background, they do not detract from the main subject and do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a large school of yellow that are predominantly yellow with black black stripes, which the description of The fish are densely in in a dense manner, which a dense group, expected. The the are other additional elements visible in the background, the do not detract from the main focus, do not contradict the description description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..33c41ce0d71ab42109eb2d08b320ab8fe11003ea
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1fe62636dd993a550539422b9015b524235cc6df097ba0f4f07a3434cc2cfc3b
+size 282181
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..54dd8ce2208fb3044d53a23e6135193ad9d09246
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a white sailboat gliding across a vast expanse of blue ocean. The boat, with its large white sails, is the main focus of the video. The sails are fully unfurled, catching the wind and propelling the boat forward. The boat is moving from left to right across the frame, leaving a trail of white foam in its wake. The ocean around the boat is a deep blue, with small waves gently breaking around the boat. The sun is shining brightly, casting a warm glow on the scene and creating a beautiful reflection on the water's surface. The overall style of the video is serene and peaceful, capturing the beauty of sailing on the open ocean."
+    ],
+    "video_ids": [
+        "JOX6uqfZNhY_17_0to131"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white sailboat with large white sails.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white sailboat with large white sails, which matches the core description. The boat is clearly visible sailing on the water with its sails fully deployed. Additional elements like the person on deck and the sparkling water do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a white sailboat with large white sails, which matches the description description. The sail is sailing visible in on the water, its sails fully unf, There elements such the water on the or the reflection water around not contradict the main but}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..62ea34a15b87470647d121b0ca6b2711402a1440
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:65997f6b5f8eb15fcc5704ec63dd4f45c7b3394a174f1975d6eaf225dce9756a
+size 237905
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..2753b607250d958bb0d4a522e4190e57ff67c610
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up shot of a car's wheel, showcasing its design and details. The wheel features a silver and black color scheme with a unique pattern of spokes and a central hub. The hub is adorned with a logo, which is a key element of the design. The wheel is captured from a low angle, emphasizing its size and design. The background is blurred, drawing focus to the wheel. The style of the video is a high-resolution, detailed shot that highlights the craftsmanship of the wheel."
+    ],
+    "video_ids": [
+        "6wkdv3B5300_80_0to130"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A car's wheel? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a close-up of a car's wheel, showcasing its design, hubcap, and tire. The wheel is clearly identifiable and central to the video, fulfilling the 'Object(s): A car's wheel' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a close-up of a car's wheel, showcasing its design and includingcap, and tire. The focus is the the as matches to the video, fulfilling the 'Object(s)' A car's wheel' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fac24c02284c33629da6d850464554378dca66f2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f45796c517dc5016947b1330d4a38d9e35e61f4e4af7ddaf586e462656685493
+size 204210
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..ee041fee5d40f4a958275dc44d4e6ea977384a04
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young boy wearing a red hoodie with bunny ears is seen riding a white unicorn with pink mane and tail. The unicorn is standing on a pink platform. The boy and the unicorn are in front of a blue elephant statue. The background features a wooden fence, trees, and a pond. The scene is colorful and playful, with a whimsical and imaginative atmosphere."
+    ],
+    "video_ids": [
+        "AtLW61IAzCw_4_0to151"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young boy wearing a red hoodie with bunny ears, a white unicorn with a pink mane and tail, and a blue elephant statue.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a young boy wearing a red hoodie with bunny ears, riding a white unicorn with a pink mane and tail. A blue elephant statue is also visible in the background. Additional elements like a sheep and another child do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a young child wearing a red hoodie with bunny ears, a a white unicorn with a pink mane and tail. Additionally blue elephant statue is also present in the background. The elements like the pink statue a elephant are not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e6c0d01afa2975631d63721c94ae98729cb1e4fa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be4ebc790aa5a88bd4fc9342145efa83021aba52ce6f4994ca2db58c2f0d84f9
+size 188405
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..3c3fcb1bec913d6a64c370e4267234f720749b08
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a polar bear in its natural habitat, navigating through a rocky landscape. The bear, with its white fur, is seen walking on all fours, its body slightly hunched over as it moves. The bear's head is turned to the side, indicating that it is aware of its surroundings. The rocky terrain is scattered with small rocks and pebbles, adding texture to the scene. In the background, a body of water can be seen, suggesting that the bear is near a water source. The overall style of the video is realistic, capturing the bear in its natural environment without any human intervention. The focus is on the bear and its interaction with the environment, providing a glimpse into the life of this majestic creature."
+    ],
+    "video_ids": [
+        "K2ROd-hKw8Y_31_93to266"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A polar bear? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a polar bear standing on rocks near water, which matches the 'Object(s)' condition. The bear's appearance, size, and behavior are consistent with that of a polar bear, even though it appears wet and is interacting with colorful objects. These additional elements do not contradict the core description of a polar bear.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a polar bear walking on a near a, which align the descriptionObject(s)' condition of The bear's white, including, and the are consistent with what of a polar bear, and though the is to, slightly not with the rocks that The additional elements do not contradict the core description of the polar bear.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2bed8e0f6ab7768d9c90995988edbad1577e39c1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:501f61fe4b02756096f5eab0276c2e702984d48f88ed83c35b14e6e3d2e29591
+size 193339
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..d104ec10b8b358ac3c0a11e3303edfc45e26ffa2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a bird's eye view of a rooftop garden in a bustling city. The garden is a vibrant oasis of greenery, with a variety of plants and trees providing a stark contrast to the urban landscape. The garden is meticulously maintained, with a well-defined layout that includes a seating area and a path leading to a small pavilion. The city skyline forms a stunning backdrop, with towering buildings and skyscrapers stretching into the distance. The video is shot from a high vantage point, providing a panoramic view of the garden and the city beyond. The style of the video is realistic, with a focus on capturing the details of the garden and the cityscape. The colors are vivid, and the lighting is natural, creating a sense of depth and perspective. The video is a testament to the beauty of urban green spaces and the importance of preserving them in the face of urbanization."
+    ],
+    "video_ids": [
+        "DKlWWmLpycc_105_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Plants and trees, seating area, path, small pavilion, city skyline, buildings, skyscrapers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing plants and trees on the rooftop, a seating area with benches, a path, a small pavilion with glass walls, a city skyline with numerous buildings and skyscrapers in the background. All specified elements are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing plants and trees, the rooftop, a seating area with benches, a path leading a small pavilion, a windows, a city skyline in various buildings, skyscrapers in the background. The these elements are present and contribute depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..df4486896ccb74038308f195bf4d411fbee7a89d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:80d10d025e8d6686fdba6176e524e55b33d83081256507842f7c4b068012ef2d
+size 237371
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..9c743c8cc259b43e43c7435974bc3697c7a36c95
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a professional audio mixing console, which is the main focus of the image. The console is large and has a variety of knobs, sliders, and buttons, indicating its complex functionality. The colors of the console are primarily black and silver, with some red and yellow accents. The console is placed on a wooden stand, which is also visible in the image. The background is blurred, but it appears to be a room with a desk and other equipment, suggesting that this is a recording studio or a similar professional setting. The style of the video is straightforward and documentary, focusing on the details of the audio equipment without any additional narrative or artistic elements."
+    ],
+    "video_ids": [
+        "2s518W5SCx0_28_0to133"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large professional audio mixing console with various knobs, sliders, and buttons.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large professional audio mixing console with multiple knobs, sliders, and buttons, which matches the description. The console has multiple sections with various controls, including labeled knobs and illuminated indicators, consistent with professional audio equipment. Additional items in the background do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close professional audio mixing console with various knobs, sliders, and buttons, which align the description provided The console is a rows with various colored, and f knobs and sliders buttons, which with a audio equipment. The elements in the background, not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c3cbcdbbc5c046a4fecd173bd21f9f4400c74dfa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1358-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59ecb4270eb8d3c066172bbec98681b0add4f6c89af73b005686b8fd29e3ca0b
+size 50419
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..698968959e679e11ddd460ac70650b31cf904318
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a cooking show segment featuring a chef preparing a roasted chicken. The chef is wearing a gray apron and is in the process of placing the chicken into a roasting pan. The pan is placed on a white cutting board, which is on a wooden countertop. The chef is using a pair of tongs to handle the chicken. There are several bowls on the countertop, containing various ingredients and sauces. The chef is also using a spoon to mix the ingredients. The overall style of the video is informative and instructional, with a focus on the cooking process. The chef is demonstrating the proper technique for preparing a roasted chicken, and the video provides a clear and detailed view of the ingredients and tools used in the process."
+    ],
+    "video_ids": [
+        "KzbUgYUodvo_13_20to153"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chef, roasting pan, chicken, tongs, bowls, spoon, various ingredients and sauces? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a chef (partially visible), a roasting pan with a chicken inside, and various ingredients and sauces (including bowls, a spoon, and seasonings). Although tongs are not visibly present, the core elements described are accurately represented in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a chef usingimially visible), a roasting pan, a chicken,, t t t and sauces insuch a with a spoon, and whatings) The theongs are not explicitly used in the action elements of in present depicted in the scene.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..38aa05b8a11eea612f70895b4273aae320756ca4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ef35d1d005618e8b619a9df6b0984cc03b790d2952be43fa64a0dcfafe9b8dd
+size 213532
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..08a9f7f4ef35129cd428a8992c8c1ffe2ffd1e7b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a grill setup outdoors, focusing on a piece of seasoned chicken being cooked over charcoal. The chicken is placed on a metal grate set inside a foil-lined tray, which rests on a red surface, possibly a table or a mat. The chicken appears to be marinated, with visible spices and herbs coating its surface. Smoke rises gently from the charcoal below, indicating the heat and cooking process. The background shows a blurred outdoor setting with greenery, suggesting a garden or backyard environment. The lighting is natural daylight, enhancing the textures and colors of the chicken and the surrounding elements. There are no significant changes or movements throughout the frames; the focus remains steadily on the grilling process, emphasizing the cooking details and the serene outdoor ambiance."
+    ],
+    "video_ids": [
+        "e10693ae4883aa2e3877b8252d23664c5b3438aa25d9e19b039c1acf76b9ea44"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Grill setup, seasoned chicken, metal grate, foil-lined tray, charcoal, spices and herbs.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a grill setup with seasoned chicken pieces on a metal grate over charcoal, contained within a foil-lined tray. Smoke is rising, indicating active grilling, and the chicken appears to be coated with spices and herbs. All elements described in the condition are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a grill setup with seasoned chicken on placed a metal grate. a. which within a foil-lined tray. The is visible, indicating the grilling, and there chicken appears to be seasoned with spices and herbs, The the described in the condition are present and accurately with the scene content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ce3fdefdde9c0ecf97e7cad95adb5f565a3aa33a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e33fbe977f0b0d60afc56d98740b4d6c405362ea9873ff96f1b0dd3383e7d0a
+size 179370
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..c01223104417d3d7cb3d6bd83e23d7b919323370
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a peaceful day in a city street. The scene is set on a tree-lined sidewalk, with a man sitting on a bench in the foreground. He is casually dressed and appears to be enjoying the day. In the background, several people are walking along the sidewalk, going about their daily routines. The buildings lining the street are tall and modern, suggesting an urban setting. The trees provide a touch of nature amidst the concrete jungle. The overall atmosphere of the video is calm and serene, with the city life unfolding in the background."
+    ],
+    "video_ids": [
+        "IddvgiKQoWg_41_0to113"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man sitting on a bench, several people walking on the sidewalk? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man sitting on a bench on the left side of the sidewalk, and multiple people are walking along the sidewalk in both directions. These elements match the 'Object(s)' condition described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts shows a man sitting on a bench, the left side of the frame. and there people walking walking on the sidewalk in the directions. The elements match the descriptionObject(s)' condition described in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6ce0642380b53eed45ed536b5af3aa99b211d623
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef9fc28ee7b395e2b525b3a835d5950f78ea2b927a738eff34695dac27e065fd
+size 425663
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..1db39e080c62826deed6eae824e848ce44830c70
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the growth of a broccoli plant in a garden. The first frame shows the plant with small green leaves and a small head of broccoli beginning to form. In the second frame, the plant has grown larger with more leaves and the head of broccoli has expanded. By the third frame, the plant has reached its full size with a large head of broccoli and a lush green canopy of leaves. The style of the video is a time-lapse, showcasing the plant's growth over time. The focus is on the broccoli plant, with the background remaining consistent throughout the video. The lighting and colors are natural, reflecting the outdoor garden setting."
+    ],
+    "video_ids": [
+        "aiZZi-Eqcf0_1_110to283"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Broccoli plant? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a broccoli plant with its characteristic green florets and large leaves, matching the description of the object. The camera movement provides a detailed view of the plant, confirming its identity without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows depicts a plant plant, its characteristic large leavesrets and leaves,. which the description of a '. The video captures and a detailed view of the plant, emphasizing its identity as any.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8f5ce354d08733e1d5f0b6c6ec6762f7f3cc2622
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a4ef5750ab3b133953e5428e05f6b3c32d33c9bfc3444ccaddcfd1dc4f793735
+size 303593
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..bfe9a7c845a0837d728271db2f62a0619632d268
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a scene in a garage where two cars are parked. The first car, a sleek black sports car, is positioned on the left side of the frame. Its shiny exterior reflects the ambient light in the garage. The second car, a dark blue sedan, is parked on the right side of the frame. Its polished surface also reflects the light in the garage. The garage itself is spacious and well-lit, with a concrete floor and a white wall in the background. The cars are stationary, suggesting a quiet moment in the garage. The video does not contain any text or additional objects. The focus is solely on the two cars and their surroundings."
+    ],
+    "video_ids": [
+        "Kyus8d9TfMs_9_92to243"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two cars - a sleek black sports car and a dark blue sedan.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two cars: a dark blue sedan in the foreground (with its taillight visible) and a sleek black sports car (Aston Martin) in the background. Both cars are clearly identifiable and match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two cars in a sleek black sedan and the background andwhich a doorillights visible) and a sleek black sports car inwithston Martin) in the background. The cars match parked visible as match the description of.}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1b7c8d80bb8e1c4123328965eba848af077bc140
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:56ea2d57aece7a40ca28f1407fbdc1b33e734f359a5ce966f284cda6219d4c5a
+size 199967
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..7cbd61862d503ef99b0f76dd69a1bbac2568ce82
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a small animated character, a girl with a pink hood, is seen interacting with a plate of food on a wooden table. The girl is holding a frying pan and appears to be cooking or preparing the food. The food on the plate is pink and resembles a dumpling or a piece of meat. The scene is set in a cozy kitchen environment with wooden furniture and a warm, inviting atmosphere. The animation style is detailed and realistic, with a focus on the character's actions and expressions. The video captures a moment of everyday life, with the character engaged in a simple yet charming activity."
+    ],
+    "video_ids": [
+        "OOgwmFJWou4_41_152to296"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Small animated girl with a pink hood, a plate of pink food resembling a dumpling or meat, a frying pan.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a small animated girl wearing a pink hood, sitting at a table with a plate of pink food that resembles a dumpling or meat. A frying pan is also visible as it is used to pour the pink food onto the plate. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a small animated girl wearing a pink hood, standing at a table with a plate of pink food that resembles a dumpling or meat. She frying pan is also present on she is being by interact a pink food onto the plate. The elements elements of in present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..264846b05b0e84f267efa7b2dae8d876e3b618b8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e79deee9e1cf3daafc6de3dc05547e6e2ef798525ce8f8757fc9b68dde1d099e
+size 136693
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..9fca35a7f3cdfde4bb35d32a7b1814d12887a2ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a large, intricately designed mosaic floor situated on a rocky beach. The mosaic features a geometric pattern with a variety of colors, including shades of red, black, and white. The floor is surrounded by large, jagged rocks, and the ocean can be seen in the background, with waves crashing against the shore. The sky is clear and blue, suggesting a sunny day. The style of the video is realistic, with a focus on the natural beauty of the location and the detailed craftsmanship of the mosaic. The camera angle is low, looking up at the mosaic, which emphasizes its size and complexity. The overall mood of the video is serene and peaceful, with the sound of the waves and the sight of the ocean creating a sense of tranquility."
+    ],
+    "video_ids": [
+        "U4WZndFzC7w_3_0to152"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Large, intricate mosaic floor, jagged rocks, ocean, crashing waves, clear blue sky? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a large, intricate mosaic floor in the foreground, jagged reddish-brown rocks along the shoreline, an ocean with crashing waves, and a clear blue sky with some clouds. All elements described in the condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a large, intricate mosaic floor with the foreground, withged rocksish-brown rocks in the shoreline, the ocean with crashing waves, and a clear blue sky. no clouds. The elements match in the condition are present and contribute represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a9aa16d7d264df99b79484dafc0906260858ae3d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c9cddf38be98f8245f40e96522f42535ff628cc717c8fa3da521264efb58425
+size 261251
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..01d8c9514ec43210cdd1241f5230842809d7711e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman in a pink floral blouse, standing in front of a colorful striped wall. She is speaking into a microphone, suggesting that she is hosting or participating in a live event or broadcast. The microphone is positioned close to her mouth, indicating that she is actively engaged in conversation or presentation. The woman's posture and expression suggest that she is confident and comfortable in her role. The colorful background adds a lively and vibrant atmosphere to the scene. The overall style of the video is casual and informal, with a focus on the woman's interaction with the audience or viewers."
+    ],
+    "video_ids": [
+        "96UHMywm92w_1054_36to173"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Woman, microphone? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman wearing a pink floral dress, and a small microphone is visible near her ear, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a woman wearing a floral floral shirt and standing there microphone microphone is positioned in her mouth, suggesting the 'Object(s)' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..aabce1de24bc070e5a482fed6a7ff41db2987f54
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1359-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a88b9fef81da6026af46d653686db66d953e9490118b7bf29962bf9e45002a60
+size 255375
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..e59a2c23f8f74d9381f9fe93f0095ea3891e30be
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a small, rectangular aquarium with a vibrant and detailed setup. The tank is filled with clear water and contains various decorative elements designed to mimic a natural underwater environment. At the center of the aquarium stands a black rock formation adorned with green moss and a small red wheel-like structure, possibly a decorative water feature or a habitat for aquatic plants. A brown stick extends from the bottom left corner, adding to the naturalistic look.  As the video progresses, several small orange fish swim gracefully around the tank. They move fluidly, exploring different areas of the aquarium, which includes colorful pebbles at the bottom and lush green aquatic plants. The plants sway gently with the movement of the water, enhancing the dynamic feel of the scene. The background remains consistently dark, which makes the bright colors of the fish and decorations stand out vividly.  Throughout the video, the camera maintains a steady position, focusing on capturing the serene and lively interaction between the fish and their environment. The overall atmosphere is tranquil"
+    ],
+    "video_ids": [
+        "e8a375c69cd58482a19df822320edf8eea3c3589f33bffc7b36d1266f2aa5100"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Black rock formation with green moss, small red wheel-like structure, brown stick, colorful pebbles, lush green aquatic plants, small orange fish? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a black rock formation with green moss, a small red wheel-like structure, a brown stick, colorful pebbles, lush green aquatic plants, and small orange fish. All these elements are clearly visible and match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a black rock formation with green moss, a small red wheel-like structure, a brown stick, colorful pebbles, lush green aquatic plants, and a orange fish. The the elements are present visible and match the description provided.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2ba4322fc07520a4ff4a87a88713700a48e4c0cf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd5c6287b63e4bea182b2776c5e47481ab79c8fd161414e355ef35e48c059428
+size 191374
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..4ec00bcbf23954e47beb4341a8eeb80aacf13603
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a herd of alpacas in a grassy field. The alpacas, with their distinctive fluffy white fur, are seen walking and grazing peacefully. The field is lush and green, providing a stark contrast to the alpacas' white fur. The alpacas are scattered throughout the field, some closer to the camera and others further away, creating a sense of depth in the image. The overall style of the video is naturalistic, capturing the alpacas in their natural habitat without any human intervention. The video does not contain any text or other objects. The focus is solely on the alpacas and their interaction with the environment."
+    ],
+    "video_ids": [
+        "bKF7rRG_tCo_4_0to135"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A herd of alpacas? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group of alpacas in a natural, arid environment, which matches the description of a 'herd of alpacas'. The animals are clearly identifiable as alpacas, and they are gathered together in a group, fulfilling the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group of animalspacas in a field setting grassid environment, which align the description of a herdherd of alpacas'. The animals are white visible as alpacas by and the are in together in a manner, fulfilling the ' of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e7953ca94d2f719f012463247bd4d8d2444efa09
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:717d1e25ae5a2d17cdbd0c711a696401ba39c8aa6ed501618dd683777f37eaaa
+size 182149
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..c3fa9705a92749b8823e43f19e66513dfc882b39
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a kitchen, standing in front of a stove. He is wearing a black shirt and has a beard. The kitchen has black cabinets and a white tile backsplash. The stove is silver and has a digital display. The man appears to be cooking or preparing a meal. The style of the video is casual and homey, capturing a moment in the man's daily life."
+    ],
+    "video_ids": [
+        "SslyC0QLSb4_1_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black shirt with a beard, standing in front of a stove.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard wearing a black shirt, standing in front of a stove in a kitchen. The core description is accurately represented, with no significant contradictions. Additional elements like kitchen cabinets, a measuring cup, and other cooking items are present but do not conflict with the primary description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a beard wearing a black shirt, standing in a of a stove. a kitchen setting The stove elements is largely represented in with the significant contradictions or The elements like the decor and a digital cup, and a kitchen-related are present but do not conflict with the main description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..86acebb24a5c3aaf3b72997c35b321ad40d9d6f2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07309656498a3e8629bff68e0f3ae772cf22ca2071d8dc67f4d7437e2adb65c4
+size 137096
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..fb1873594ec3725085a2c714c9483bfff664ad13
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a meal being served on a white plate. The meal consists of a bowl of soup, a bowl of sauce, and a serving of white rice with pieces of chicken on top. The soup is a light broth with green herbs, and the sauce is a dark brown color. The chicken is white and appears to be cooked. The plate is placed on a table, and the meal is being served to a person who is not visible in the video. The style of the video is a simple, straightforward depiction of a meal being served, with no additional context or embellishments."
+    ],
+    "video_ids": [
+        "UzOcnQSJnSA_27_0to152"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white plate, a bowl of light broth soup with green herbs, a bowl of dark brown sauce, a serving of white rice with pieces of white cooked chicken on top.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white plate with a serving of white rice topped with pieces of white cooked chicken, a bowl of light broth soup with green herbs, and a bowl of dark brown sauce. All elements described in the condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white plate with a bowl of white rice topped with pieces of white cooked chicken. a bowl of light broth soup with green herbs, and a bowl of dark brown sauce. The elements match in the condition are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5b2bfd2aaaa8f8aa0534b255c5c3545ef84c6c43
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:390281d8f6924e100c75e862fc775f514046c2fe0f42666f1c9b92f74a8a12a5
+size 59649
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..60058f5651b98d353efe01069850f490a0c7a9f8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red sports car parked on the side of a road. The car is sleek and shiny, with a black spoiler on the back and a black front bumper. The car's wheels are black with silver rims. The car is parked on a concrete curb, and the road behind it is clear. The sky is blue and clear, suggesting a sunny day. The car is the main focus of the video, and there are no other objects or people in the frame. The style of the video is a straightforward, clear shot of the car, with no additional action or movement."
+    ],
+    "video_ids": [
+        "Fj8bUe8--gI_1_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red sports car, which matches the description. The car is clearly visible, and while there are additional elements in the background, they do not contradict the core description of a red sports car.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a red sports car, which matches the description of The car is shown visible, with its there are some elements like the background, such do not conflict the core description of the red sports car.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5a0bafc18b239baae52ccbc92cce0e9e439ff65f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2db8a1b4fa14e9709733ea104169073dc6727f0d6582d38b9d584df6efd984d2
+size 179206
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..2bf754eade5d0c9c065e2a64fa3b732cc4feeaf0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse of a car's interior, showcasing the progression of the car's interior from day to night. The car is a luxury vehicle with a sleek design, featuring a wooden dashboard and a red and black steering wheel. The car's interior is well-lit, with the sun shining through the windows during the day, and the interior lights illuminating the space at night. The car's seats are black leather, and the dashboard is made of wood, adding a touch of elegance to the vehicle. The car's steering wheel is red and black, and the dashboard features a red and black logo. The car's interior is well-maintained, with no visible damage or wear. The car's interior is clean and well-organized, with no clutter or debris. The car's interior is well-lit, with the sun shining through the windows during the day, and the interior lights illuminating the space at night. The car's seats are black leather, and the dashboard is made of wood, adding a touch of elegance to the vehicle. The car's steering wheel is red and black, and the dashboard features a red and black logo. The car's interior is well-maintained, with no visible damage or wear. The car's interior is clean and well-organized, with no clutter or debris. The car's interior is well-lit, with the sun shining through the windows"
+    ],
+    "video_ids": [
+        "sDACV60VyQI_18_19to196"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Luxury car with wooden dashboard, red and black steering wheel, black leather seats, and a red and black logo on the dashboard.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows the interior of a luxury car with a wooden dashboard, black leather seats, and a red and black logo on the dashboard. The steering wheel is black, and while it's not explicitly red, the overall aesthetic matches the description. The presence of additional elements like the 'Vehicle Virgins' watermark does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts the interior of a car car with a wooden dashboard, a leather seats, and a red and black steering on the dashboard. The steering wheel is also with but there it does not explicitly mentioned, the video color and the description. The presence of additional elements like the sunA'alins' logo does not conflict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..27e7afde97ad3382ffe4a697d435c17bcc4dcefc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b16d8bfb8d5f15544e2da81e88fb583bad92a28c4f793f2ad1644bc7aeded411
+size 180282
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..8e20b5c5b25c20c878301df326078915d7c6e7f4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures an underwater scene teeming with marine life and vibrant coral formations. The camera glides smoothly over a lush coral reef, showcasing a variety of corals in different shapes and sizes, ranging from branching to rounded forms. The colors of the corals vary from deep greens to lighter hues, creating a rich tapestry beneath the water's surface.  As the camera moves forward, it reveals a school of fish swimming gracefully among the corals. These fish, characterized by their dark bodies with white stripes, glide effortlessly through the water, adding a dynamic element to the serene underwater landscape. The fish appear to be exploring the reef, occasionally darting between the branches and crevices of the coral formations.  The camera continues its journey, capturing the intricate details of the coral structures and the fluid motion of the fish. The interplay of light and shadow enhances the visual appeal, highlighting the textures and contours of both the corals and the fish. The overall atmosphere is one of tranquility and natural"
+    ],
+    "video_ids": [
+        "1e621229ce3128cb4de65fdb597128e0176a92eab82ffed9af13bd16bcf49fbc"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Corals, fish? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows corals and fish as requested. The underwater environment is filled with various types of coral formations, and a fish is clearly visible swimming among them. The presence of other marine life and coral structures does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful aals, does. it. The frames scene depicted depicted with vibrant cor of cor,, which the fish can visible visible swimming among the. The vibrant of the elements life or the types align not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e83a3e9db4b5a4c5a6181d5e25805bac2c4f8f86
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4426b14f607802eb2dce3afe100405a176c82bd1e1dae3a6ed8f47665ffbef82
+size 241062
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..37565daf25db22957b146d2f1221c53cbcda5bf0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with glasses, who is the main subject. He is wearing a black shirt and is standing in front of a lush green background filled with plants and trees. The man appears to be speaking, as suggested by his open mouth and focused gaze. The overall style of the video is natural and organic, with the man's attire contrasting against the vibrant greenery. The video captures a moment of interaction between the man and his surroundings, creating a sense of connection with nature."
+    ],
+    "video_ids": [
+        "45gBLiC1r4s_13_0to122"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with glasses, wearing a black shirt, speaking? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with glasses and a black shirt who is speaking. The background includes plants and people, but these do not contradict the core description of the man speaking. The visual elements are consistent with the given condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing glasses wearing a black shirt, appears speaking. The background is green, green, but these elements not contradict the core description of the man's. The presence elements in consistent with the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b841f4775ae8ddd8fc21b282c9c02de40a2fad16
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-136-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eca4c151b2c5535f61b26df4651fe17fc3e1cce4986400101790dc4d1938c77f
+size 97020
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..8f899a35b65542917c30eda46897a8e791e4810e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up of a bowl of food, showcasing a variety of ingredients. The bowl is filled with a mix of meat, vegetables, and noodles. The meat appears to be beef, sliced into thin pieces. The vegetables include green beans, carrots, and onions, adding a pop of color to the dish. The noodles are thin and appear to be cooked to perfection. The dish is presented in a black bowl, which contrasts nicely with the vibrant colors of the ingredients. The style of the video is simple and straightforward, focusing solely on the food without any additional elements or distractions. The video is likely intended to showcase the dish's ingredients and presentation, possibly for a recipe or food review."
+    ],
+    "video_ids": [
+        "8soqNhejOJs_28_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Meat (beef), vegetables (green beans, carrots, onions), noodles? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a bowl containing sliced beef, green beans, carrots, onions, and noodles, which matches the specified objects. The visual presentation aligns with the description without any contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a dish of a meat, green beans, carrots, and, and noodles. which align the description objects. The presence elements ofs with the description of any contradictions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1f9608efa9dc790aba5487a093175f7e19c8cfe1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b4b99aa54afe191d66289e65f1a95e7ec5abccb726136546ff01a55bdea80ce
+size 114008
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..14aab4816face16fa13df074b961fa6edea69de8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated sequence featuring three dolls with different hair colors and styles. The dolls are positioned close to each other, facing each other, and appear to be engaged in a conversation. The background is a simple, nondescript room with a white brick wall and a small, purple flower in a vase on a shelf. The dolls are dressed in colorful clothing, with one wearing a pink top and the other in a purple top. The animation style is smooth and detailed, with realistic textures and shading on the dolls' hair and clothing. The overall mood of the video is friendly and cheerful, suggesting a light-hearted interaction between the dolls."
+    ],
+    "video_ids": [
+        "CRZCOi8R5Ts_60_0to207"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three dolls with different hair colors and styles, dressed in colorful clothing.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three dolls with distinct hair colors and styles: one with blonde hair, one with long orange hair, and one with red curly hair. Each doll is dressed in colorful clothing\u2014an orange top with a purple collar, a floral-patterned top, and a partially visible outfit. These elements match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two dolls with different hair colors and styles, one with blonde hair, one with red red hair, and one with red hair hair. They doll is dressed in a clothing, orange dress, a pink collar, a pink dressed top, and a pink visible outfit with The elements align the description of,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b6713b24dba8e5a3f9d20427da3a31ab53ead02c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:836675fe1d24198bdaa61edee8d6a880ccb98d71fbb809ae0b261857d56f3b3e
+size 202119
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..cb2c5a577339d4417de67e3eca3874647a4f290b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a lively street scene at night. A woman, dressed in a green jacket, is seen holding a microphone, presumably conducting an interview. She is standing in front of a man who is wearing a white shirt with a green logo. The man appears to be the subject of the interview. The background is bustling with people, adding to the vibrant atmosphere of the scene. The video seems to be a news report or a street interview, capturing a moment in the life of the city."
+    ],
+    "video_ids": [
+        "b2rCuQVAT1M_5_376to513"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a green jacket holding a microphone and a man wearing a white shirt with a green logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a green jacket holding a microphone and a man wearing a white shirt with a green logo (the letter 'P' on his shirt). These elements match the description provided. The background crowd and setting do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a green jacket holding a microphone and a man wearing a white shirt with a green logo.which letter 'F' is the shirt). The elements match the description provided, The background is and the suggest not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..51f4e8198bce290da832a86c55d126b7e82472f7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:651df10f190b6e5cd59fc54e40e81707633f5a3b9805267e38a13957abb3bd41
+size 201611
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..513179dba850954eca03aa2beb2d75f5ddc4d715
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two scientists in white lab coats and blue goggles are engaged in a discussion. They are standing in a lab, surrounded by various pieces of equipment. The scientist on the left is holding a tablet, while the one on the right is holding a clipboard. They are both looking at the tablet, which is displaying some data. The scientist on the left is pointing at the tablet, indicating a specific piece of information. The scientist on the right is nodding, indicating that he understands or agrees with what the other is saying. The lab is well-lit, and the equipment is neatly arranged. The scientists appear to be focused on their work, and the atmosphere is one of concentration and collaboration."
+    ],
+    "video_ids": [
+        "SpKGdt9BHo0_12_0to156"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two scientists in white lab coats and blue goggles.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men dressed in white lab coats and wearing blue goggles, consistent with the 'Object(s)' condition. Although other elements like a tablet, background equipment, and a passing person are present, they do not contradict the core description of the two scientists.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two individuals wearing in white lab coats and wearing blue goggles, which with the descriptionObject(s)' condition. They there elements like laboratory laboratory and laboratory lab, and additional third person are present, they do not contradict the core description and the two scientists in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e51e277c37596029ee90937a9a3b521918b48bc7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:319a470a22ced410160ae5d537a7c739c3bd42440183cc09f4d0971173955a8e
+size 135951
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..115abc8643ae2ae3a5acf5bf187da3c9561e5d6e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a blue sports car parked on a brick driveway in front of a house. The car is sleek and modern, with black rims and orange accents on the wheels. The house has a red brick facade and a white porch. The car is the main focus of the video, and it is shown from different angles as it moves along the driveway. The video has a smooth and polished look, with a focus on the car's design and the house's architecture. The car's movement suggests that it is being driven, and the house's location suggests that it is in a suburban area. The video does not contain any text or additional objects."
+    ],
+    "video_ids": [
+        "RFKM2xN-wRc_8_299to424"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue sports car, a house? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a blue sports car parked in front of a house, matching the 'Object(s)' condition. The car is prominently featured, and the house is visible in the background. Additional elements like trees and a driveway are present but do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a blue sports car parked in front of a house. which the descriptionObject(s)' condition. The car is prominently featured, and the house is visible in the background, There elements like the and a driveway are present but do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bb4f06c9c67fecfa5da13c717346a7dc9bd41634
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce8fbbb9d6b6152aba2167011e3812ce0a5f1348498aca945b87590fd9ba930a
+size 200282
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..3984fb20444fc05371942535382fb175c8578e38
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a yellow shirt, who appears to be in a state of surprise or shock. He is looking off to the side, with his mouth slightly open, suggesting that he has just seen or heard something unexpected. The background is filled with colorful abstract art, which adds a vibrant and lively atmosphere to the scene. The man's expression and the abstract art in the background create a dynamic and engaging visual narrative."
+    ],
+    "video_ids": [
+        "Eg6MEWVGdgw_19_39to256"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a yellow shirt, showing surprise or shock with his mouth slightly open.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a yellow shirt who appears to be reacting with surprise or shock, as indicated by his wide eyes and slightly open mouth. His facial expression and body language align with the described condition, even though the background contains additional decorative elements that do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a yellow shirt with appears to be expressing with surprise or shock, as indicated by his slightly eyes and slightly open mouth. The facial expression align body language align with the description ' of making though the background is additional elements elements that are not conflict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..835f8494394cafdbfde5332e5dc1babd26d7521e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d78cdde11981eb91009f0693427eff183018c5901b754825ae05a4d25465f171
+size 143036
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..338cf8934123e68e443d46c0aa725a59f2d6c95a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a culinary journey featuring a bowl of tomato soup. The first frame shows a close-up of the soup, with a spoonful of butter being added to the bowl. The second frame captures the butter melting into the soup, creating a rich and creamy texture. The third frame reveals the final product, a bowl of delicious tomato soup with a dollop of butter on top. The style of the video is simple yet effective, focusing on the preparation and presentation of the soup. The use of close-ups and slow motion shots highlights the details of the ingredients and the cooking process, making the viewer feel as if they are in the kitchen with the chef. The overall effect is a mouth-watering visual feast that showcases the beauty of simple, home-cooked meals."
+    ],
+    "video_ids": [
+        "vzKr78XucbE_10_0to155"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Bowl of tomato soup, spoonful of butter? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a bowl of tomato soup with a spoonful of butter on top, which matches the description. The surrounding tomatoes and bamboo mat are additional elements that do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a bowl of tomato soup and a spoonful of butter being top. which matches the description of The sequence elements and the mat in additional elements but do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8300390eb8f54308e4a71c096d04ffad46cf86cd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bdfd26dd13a3a91de1a52728a4870ace4f0564b5edb6d8021529861663295a69
+size 184658
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..efd6afc015233d1fbacf456a638547d08d3810f5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video begins with a close-up shot of a dark-colored beetle resting on a sandy surface. The beetle is positioned centrally in the frame, with its body oriented diagonally across the sandy ground. Its legs are spread out, and it appears to be stationary. The sand around the beetle is fine-grained and light brown, indicating a dry environment. The beetle's shell is shiny and reflective, suggesting it might be wet or coated with a glossy substance. The background is a uniform sandy texture, emphasizing the isolation of the beetle.  As the video progresses, there is no significant movement from the beetle itself. However, the scene transitions to show the same sandy area, but now the beetle is no longer visible. Instead, a small, reddish-brown insect, possibly an ant, is seen moving across the sand. This insect is much smaller than the beetle and has a more elongated body shape. It appears to be crawling quickly, leaving a trail of disturbed sand behind it. The camera remains"
+    ],
+    "video_ids": [
+        "f39a1c4a0f8591b5e84428f7d872565efe313a7769e5e15d422fc01d0ecde776"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A dark-colored beetle and a small, reddish-brown insect (possibly an ant).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dark-colored beetle on sand, and a small, reddish-brown insect (likely an ant) briefly interacts with it. The core objects described are present, and no major contradictions exist.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dark-colored beetle, a, which a small, reddish-brown insect,possibly an ant) appears appears with the. The presence elements described in present, and the additional contradictions are.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6cc3a84bc0e4260d776cafa3c21157c31f6a9c7a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1360-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f8abce767f770b3fead478d26cf33922e0b7e9f1c00da928eb659849acf0350
+size 236470
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..44e62970df4e6b561f715ded45a6c10ba61feec3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a character in a fantasy setting, standing in a snowy landscape with mountains in the background. The character is wearing a dark, armored suit with spikes and a helmet, giving off a menacing and powerful aura. The armor is detailed and intricate, suggesting a high level of craftsmanship. The character is standing with their arms outstretched, as if commanding or controlling something. The overall style of the video is realistic with a touch of fantasy, and the lighting is soft, casting a warm glow on the snowy landscape. The character's pose and the surrounding environment suggest a sense of adventure and exploration."
+    ],
+    "video_ids": [
+        "h6p7da3t1do_1_0to170"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A character in a dark, spiked, and detailed armored suit with a helmet.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a character wearing a dark, spiked, and detailed armored suit with a helmet, which matches the description. The armor is intricately designed with sharp, angular spikes and a helmet that covers the head, fitting the specified criteria.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a character wearing a dark, spiked, and detailed armored suit with a helmet. which align the description provided The armor is adornedately designed with spikes edges angular spikes and a helmet that covers the head, fitting the ' '.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7f465b23e659eba1c75fc31f72dba5f69b1040f3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a55c61f56e16f18a4898a6c495ef56e4152772cce8c369b4cb923fbc800b53ef
+size 169984
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..99d307bb8afdba8fafbe4569a4ff9f82d5d9bd1d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring an older woman in a factory setting. She is wearing glasses and a brown shirt with a white apron. The woman is speaking and appears to be explaining something. In the background, there are other workers and machinery, indicating an industrial environment. The style of the video is informative, with a focus on the woman's speech and the context of her work. The video likely includes a voiceover or subtitles to provide additional information about the woman and her role in the factory."
+    ],
+    "video_ids": [
+        "QJMYG5HsM5o_83_0to119"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An older woman wearing glasses, a brown shirt, and a white apron, speaking and explaining something.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an older woman wearing glasses, a brown shirt, and a white apron with a pattern. She is actively speaking and gesturing as if explaining something. The background elements, such as tables and other people, do not contradict the core description and are consistent with a workshop or factory setting.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video features an older woman wearing glasses, a brown shirt, and a gray apron. a name that She is speaking speaking and appearsuring, if explaining something. The background is, such as the and chairs figures, do not contradict the core description and are acceptable with a setting or educational setting,}<|im_end|>\nGuidId",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..049e8effd7231207103fa3701a0f103572f3fa30
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a523c484e75c7ec648402d8a28ad7c82aef8ed1c108bab5513d51db1afdfa55
+size 155251
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..c63899d5e924ded9fca25be664a4e6b53f0c17d2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a whimsical scene of an alien spaceship hovering over a suburban neighborhood at night. The spaceship, with its blue and yellow lights, is the central focus of the image, appearing to be in the process of landing. The neighborhood is composed of three houses, each with a unique architectural style, and a tree in the foreground. The houses are illuminated by the spaceship's lights, creating a contrast against the dark night sky. The overall style of the video is cartoonish and playful, with a vibrant color palette that adds to the fantastical nature of the scene."
+    ],
+    "video_ids": [
+        "hdgXvXNGqw0_210_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An alien spaceship, three houses, and a tree? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows an alien spaceship (a UFO with glowing lights) hovering over a suburban neighborhood. There are three distinct houses visible: one on the left, one in the center, and one partially visible on the right. Additionally, a tree is present on the right side of the scene. All core elements from the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts an object spaceship hoveringa car-like a lights and flying above a residential neighborhood. There are three houses houses visible in one with the left, one in the middle, and one on visible on the right. Additionally, there tree is present in the right side of the image. The elements elements of the description are present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..612a689782239b20ab4d62a9c7d723c772851cff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bddefd2d79d85aaf79cb9d84117ad1d46db87680d0a6e5e7f10989ab4996319b
+size 152307
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..e4c477370946e5cc9bfd4c9baa1f8ffc325a2ae9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic scene on a dirt road. An orange Ford truck is seen driving down the road, its vibrant color contrasting with the earthy tones of the surroundings. The truck is in motion, suggesting a sense of speed and adventure. The road itself is unpaved, adding a rugged charm to the scene.  In the foreground, a white truck is parked, its shiny exterior reflecting the bright sunlight. The white truck is stationary, providing a contrast to the moving orange truck. The two trucks are positioned in such a way that they appear to be moving in opposite directions, adding a sense of depth and perspective to the scene.  The sky above is a clear blue, suggesting a bright and sunny day. The sunlight illuminates the scene, casting long shadows and highlighting the textures of the dirt road and the trucks. The overall style of the video is realistic, capturing the details of the trucks and the road with precision. The composition of the scene, with the two trucks in the foreground and the open road stretching out behind them, creates a sense of anticipation and adventure."
+    ],
+    "video_ids": [
+        "aGoNhITXEu8_65_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An orange Ford truck and a white truck? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows an orange Ford truck approaching and stopping next to a white truck, fulfilling the 'Object(s)' condition as described. The presence of additional elements like the desert background or the 'TFLTRUCK' logo does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows an orange Ford truck and from a behind to a white truck. which the 'Object(s)' condition. described. The presence of additional elements like the road landscape and the treesb''UCK' text does not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a87d52d2ba5b6eed5bdcc67da170b6037ee60abc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:adaf8e04a2dbdaf92e3fbeb5be5a12fd5272aac89ad0bb982c82e394331533a0
+size 380922
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..9ae4e17d39c70fde0c7c81a778e47bb12d1db6ba
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a plaid shirt standing in front of a white background with three monkey faces. The man is wearing glasses and appears to be speaking. The monkey faces are arranged in a row, each with a different expression and color. The first monkey is gray, the second is red, and the third is brown. The man is positioned between the second and third monkey faces. The overall style of the video is simple and straightforward, with a focus on the man and the monkey faces."
+    ],
+    "video_ids": [
+        "c_jyHp3bmEw_80_0to186"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a plaid shirt, three monkey faces.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man wearing a plaid shirt and three distinct monkey faces (gorilla, orangutan, and chimpanzee) that appear sequentially on the left side of the screen. These elements are clearly visible and match the described 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a plaid shirt, three monkey monkey faces areoneilla, orangutan, and chimpanzee) super are to in the screen side of the frame. The elements match consistent visible and match the description 'Object(s)' condition without any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a356f1e0cd87152f2cac882de3917462c51e81a5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:874468162e994d39771c1cfe672908fafee7804d8365250765b9ccd177045350
+size 184555
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..04064d140a4184dbedffb721495b0795ef26c9b4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a gray shirt and sunglasses standing in front of an open car hood. He appears to be speaking or gesturing towards the engine. The car is parked in a driveway with a wooden fence in the background. The man's style is casual and he seems to be explaining something about the car. The setting is a residential area with greenery and a blue bucket visible in the background. The video is likely educational or informative, possibly related to car maintenance or repair."
+    ],
+    "video_ids": [
+        "S3xye5akLL0_21_0to177"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a gray shirt and sunglasses, an open car hood, a parked car, and a blue bucket.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a man in a gray shirt and sunglasses standing in front of an open car hood. A parked car is visible in the background, and a blue bucket can be seen on the ground to the left side of the frame. Additional elements like a superimposed car image and a fan do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as depicting showing a man wearing a gray shirt and sunglasses standing next front of a open car hood. The parked car is visible, the background, and a blue bucket is be seen to the grass to the left of of the frame. The elements like the woodenimposed logo logo in a green in not contradict the core description and}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c0ccab1b7968c45fcb9a2dfd4516de061e1fedd7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07063d6e6366ffcf400366a0553eea63cd726d2ecc7a26aa01f731fda84afc96
+size 301475
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..3ab582276be7587a7bfb77e2b26442a11b205354
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen sitting at a red table, holding a tablet in her hands. She is wearing glasses and a black and white patterned blouse. The setting appears to be a television studio, as indicated by the presence of a blue couch and a man in the background. The woman seems to be engaged in a conversation or presentation, as suggested by her focused gaze on the tablet. The overall style of the video suggests a professional or informative tone, possibly related to a news or talk show segment."
+    ],
+    "video_ids": [
+        "0g2gdC5Ei0A_35_21to161"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a tablet, glasses, a black and white patterned blouse, a blue couch, and a man.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a woman wearing glasses and a black and white patterned blouse, holding a tablet, seated at a table. A blue couch is visible in the background, and a man is present in the scene, either seated on the couch or facing the woman. All core elements from the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows ful a woman wearing glasses and a black and white patterned blouse, holding a tablet. and at a red. The blue couch is visible in the background, and there man is not in the scene, though partially or the couch or standing the woman. The elements elements of the description are present, contradiction.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..274c7ff14e8d60d29127f94d26a89c957fad68e6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:20ae3de745fbbd2d0ec525ffe5d544232aa7d013c7a9b1cc373a5c43c09fcbac
+size 177116
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..bffb85b7af88e5fbfb148ff312c745da32545bfc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated depiction of a medieval king standing in a grand hall. The king is dressed in a red and gold armor, complete with a crown on his head. He is standing on a checkered floor, with a large archway in the background. The king's expression is serious, suggesting a moment of decision or contemplation. The overall style of the video is realistic, with attention to detail in the armor and the grand hall. The lighting and shadows add depth to the scene, creating a sense of realism. The video captures the majesty and power of the king, as well as the grandeur of the setting."
+    ],
+    "video_ids": [
+        "J-eN4QSMz0o_17_0to120"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A medieval king wearing red and gold armor with a crown.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a medieval king wearing red and gold armor with a crown, matching the core description. The armor is detailed with red fabric and gold accents, and the crown is clearly visible on his head. Additional elements like the background setting and other figures do not contradict the primary description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a character king wearing red and gold armor with a crown. which the description description. The attire is detailed with red and and gold accents, and the king is clearly visible, the head. The elements such the check architecture do the characters are not contradict the main description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..236dd848551699e510da0e21cbbbbd86dd40641c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1361-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:69c8709ee7d2075e23e134b3985f4e81574691a4030926bcbac4e90fa3d8b39e
+size 89556
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..0ae261f60f509b90feeed8cbfb47129a90030e29
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of an elephant's face, focusing on its wet skin and expressive eyes. The elephant appears to be enjoying a refreshing bath, as water droplets are seen splashing around its trunk and head. The background is a lush green, indicating a natural setting, possibly a forest or a wildlife sanctuary. The camera remains steady throughout the sequence, maintaining a consistent focus on the elephant's face, highlighting the texture of its skin and the gentle movement of the water droplets. The overall atmosphere conveys a sense of tranquility and the beauty of nature."
+    ],
+    "video_ids": [
+        "aa1fdd6aa0c5703f108ba029d64afe57388a8c387be1c2d1547aefc062595ff7"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An elephant, water droplets? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows an elephant with visible water droplets spraying onto its face, matching the described objects. The elephant's wet skin and the motion of water droplets are consistent with the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts an elephant, water water droplets on from its face. which the ' '. The presence is trunk skin and the presence of the droplets are prominent with the ' of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ff3ea1c62e59beaabb89d7c1ffaef0f442204d10
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b1511fdb469b12af17b742753bb72886de7c59031df1c7a2207414e9ce5f7af
+size 368024
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..59a13848c14645ff136cd089dcd8b6faad55c2c2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up shot of a motorcycle tire, showcasing its tread pattern and design. The tire is black with a series of grooves and sipes that are typical for motorcycle tires, designed to provide grip and traction on various road surfaces. The motorcycle is parked in a garage or workshop, with a concrete floor and metal beams visible in the background. The lighting is bright, highlighting the details of the tire and the surrounding environment. The style of the video is straightforward and informative, focusing on the tire's features without any additional context or narrative."
+    ],
+    "video_ids": [
+        "4SXKDNAcgbQ_11_0to167"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A motorcycle tire with a black color, featuring a series of grooves and sipes.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a motorcycle tire that is black in color and clearly shows a detailed pattern of grooves and sipes on its tread. These elements are consistent with the description provided, and no conflicting visual elements are present that would contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a motorcycle tire with is black in color, displays displays a series pattern of grooves and sipes. its surface. The features are consistent with the description provided, indicating there conflicting elements elements are present in would detr the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0e5597d014ab183a175b18a16eb83af3475dd3db
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1c0d887904313dc6113cee27773e7a038d219442e86c455077d1d0039008b1d
+size 200776
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..76500e27d04cbfe7d2895bfb2eda28f899c9cf95
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated scene featuring five children gathered around a baby in a crib. The children are diverse in appearance, with different hair colors and styles, and they are all smiling and looking at the baby with affection. The crib is white with a red and yellow striped mattress, and the baby is lying down, looking peaceful. The room has a pink and white striped wallpaper, and there is a window with a white frame in the background. The overall style of the video is colorful and cheerful, with a focus on the children's expressions and the baby's calm demeanor."
+    ],
+    "video_ids": [
+        "HN5ibJuL9xQ_9_0to102"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Five children and a baby in a white crib with a red and yellow striped mattress. The children are diverse in appearance, with different hair colors and styles, and they are all smiling and looking at the baby with affection.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows five children surrounding a white crib with a red and yellow striped mattress, and a baby inside. The children are diverse in appearance, with different hair colors and styles, and they are all smiling and looking at the baby with affection. The scene matches the description without significant contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts five children and a baby crib with a red and yellow striped mattress, where a baby lying the The children are diverse in appearance, with different hair colors and styles, and they are all smiling and looking at the baby with affection. The scene matches the description provided any contradictions or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1d6712df2f24baae9ac9dd19c46a213de759e60d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9abaf34def06c606e1138d4e4c0b52998f6d17df20e4929784d2bf6128698892
+size 110927
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..65db147b974adfdaa4b810efe50d3c0fef3ab750
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young man and woman are seated at a table, engaged in a playful activity. The man is holding a small potato in his hand, while the woman is writing in a notebook. The table is adorned with a vase filled with colorful pencils, adding a vibrant touch to the scene. The backdrop features a bookshelf filled with books, suggesting a studious or intellectual setting. The overall style of the video is casual and light-hearted, capturing a moment of fun and creativity between the two individuals."
+    ],
+    "video_ids": [
+        "RJwB-agReSU_17_88to231"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man holding a small potato, a woman writing in a notebook? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man holding a small potato (which appears to be a cookie or similar object) and a woman writing in a notebook. These core elements are present and consistent with the description, even though there are additional decorative and playful elements in the scene (e.g., crayons, green antennae, colorful background).\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man holding a small potato andobject could to be a toy or a object) and a woman writing in a notebook. The elements elements match present and match with the description. even though the are additional elements elements background elements like the scene,such.g., aons, aeryae, and background). The}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..420239cbb48d2c09f13fe8ee2e97f8be9e67fccb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bbb363db8c8d72c7f87f879ce770cffbb7a5b2ef427c86e7e68dc5285d6a432e
+size 120430
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..1d59a3bf3cf0c44c3ee51411afb1aeff2b30e725
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a Christmas tree adorned with various ornaments. The tree is decorated with a mix of traditional and unique ornaments, including a white bird, a red ball, and a gold ribbon. The bird, which is the focal point of the video, is perched on a branch, adding a touch of whimsy to the festive scene. The red ball and gold ribbon add a pop of color and a touch of elegance to the tree. The tree is illuminated by twinkling lights, creating a warm and inviting atmosphere. The video is shot in a way that allows the viewer to appreciate the details of the ornaments and the overall beauty of the Christmas tree."
+    ],
+    "video_ids": [
+        "sxqMvlSkxr0_23_314to443"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Christmas tree, white bird, red ball, gold ribbon? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a decorated Christmas tree with a white bird ornament, a red ball ornament, and a gold ribbon. These elements are prominently visible and match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a Christmas Christmas tree with various white bird ornament, a red ball,, and a gold ribbon. The elements match consistent featured and match the description provided.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..91b01da4e4c8ddd9c31530fe5c705b05b0881512
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d0f107cba93e9858997dc4799e8b89e1108171163bae51ddfb6dcb42ac6012c7
+size 70458
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..4cac05e4365b8069ba0a147b5eed2ce4d2d4e5cd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and glasses, wearing a red shirt with a yellow badge. He is standing in a cluttered room filled with various items, including books, boxes, and bottles. The man appears to be in the middle of a conversation or explanation, as he gestures with his hands. The room has a rustic and somewhat disorganized appearance, with items scattered around and stacked on shelves. The lighting in the room is dim, creating a somewhat moody atmosphere. The man's facial expression suggests that he is serious or focused on the topic at hand."
+    ],
+    "video_ids": [
+        "B7lgo1LbpCs_30_0to101"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and glasses, wearing a red shirt with a yellow badge.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The man in the video has a beard, is wearing a red shirt, and has a yellow badge (with 'B' and 'BROON' visible) on his shirt. He also has glasses hanging from his shirt. These elements match the description provided in the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video in the video has a beard and is wearing glasses red shirt, and has a yellow badge onwhich aUPS' on 'S'OK' written) on his shirt. The is wears glasses, on his shirt. The elements match the description provided, the question.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..791963694764288898262a0de768af3916a92ae1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2857be576f5c5848d7c1e4c5cfb7dd44ee79a387096952b8d551dbab8aea23eb
+size 170844
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..23855c9a9fb7a1757d18072b043d70feef995cb6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a journey on a ferry boat, with a focus on the vehicles and the surrounding landscape. The first frame shows a white RV parked on the ferry, with a clear blue sky and a mountain in the background. The second frame shows the RV driving off the ferry onto a road, with the same mountain in the background. The third frame shows the RV driving down the road, with the mountain still visible in the distance. The style of the video is a combination of travel and nature, with a focus on the journey and the beautiful landscape."
+    ],
+    "video_ids": [
+        "xao70U7oKYM_53_0to130"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white RV? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a white RV parked on the left side of the road, which matches the 'Object(s)' condition. The RV is prominently visible and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a white RV in on a side side of the frame in which ful the descriptionObject(s)' condition. The RV is the featured in is in the description.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..742de2c62025c04f7dfef9c2042b2b0cb057e19c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:565d4a603f08729b3d23dff1e4ec6b6222fa8598c73c8b8969a93cbe6fbb4de4
+size 269855
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..4837b2844c6e6cb48a34c79a05e4b3488b6c8b0e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a journey on a desert road. A yellow sports car is the main subject, driving down the road, leaving a trail of dust behind. The road is surrounded by a barren landscape, with a few scattered trees and power lines. The sky above is clear and blue, suggesting a sunny day. The car's speed and the dust trail indicate movement, while the stillness of the surrounding environment contrasts with this motion. The overall style of the video is a blend of action and stillness, capturing the essence of a road trip in a desert setting."
+    ],
+    "video_ids": [
+        "2yD3bdFApxc_18_0to173"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A yellow sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a yellow sports car, specifically a Chevrolet Corvette, driving down a road. The car is the central focus and matches the description of a yellow sports car.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a yellow sports car driving which a model Corvette, driving on a road. The car is the central object of matches the description of a yellow sports car.}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7598570a3fb41c8af83a0a129bda76be08ea6ffc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1362-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee6793e8928b7b5bbbdb195bd8083d01457dc3227a8c1dc289d7a899f4cdeee5
+size 202681
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..ba35a3d43b7d28833fe07faa0431e630cc557099
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up of a black and yellow cobra snake in its natural habitat. The snake is seen in three different positions, each showcasing its unique features. In the first frame, the snake is coiled up, its body tightly wound as it prepares to strike. The second frame shows the snake in a more relaxed state, its body partially uncoiled as it moves through the dirt. The third frame captures the snake in a striking pose, its body fully extended as it prepares to strike. The snake's black and yellow scales are clearly visible, adding to the overall intensity of the scene. The background is a simple dirt ground, providing a stark contrast to the snake's vibrant colors. The video is a fascinating glimpse into the life of this majestic creature."
+    ],
+    "video_ids": [
+        "EvV_XrYG1xw_54_20to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black and yellow cobra snake? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a cobra snake with distinct black and yellow coloring, particularly visible on its hood and neck. The snake is in a raised, defensive posture, which is characteristic of cobras. The background and surrounding elements do not contradict the description of the snake itself.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a blackobra snake with a black and yellow color, which noticeable in its head and body. The snake's depicted a striking position defensive posture, which is characteristic of cobras. The presence and the environment, not contradict the description of the snake being.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5ff8df207a4310ca278bf30b744c65a888c33174
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c0e0be7366327738b99bcc00eabc5c300d83b6ee2d222ba71cfc0a7ddb8cc4b
+size 349406
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..d49f792e0f739834228adfbbcab992d569948d0e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a colorful and playful animation featuring a group of stuffed animals on a bed. The bed is adorned with a blue comforter and a pink headboard. The stuffed animals include a lion, a dinosaur, an elephant, and a duck. The lion is on the left side of the bed, the dinosaur is in the middle, the elephant is on the right side, and the duck is at the foot of the bed. The stuffed animals are arranged in a line, facing the viewer. The background is a simple pink wall with a gold lamp on the right side. The overall style of the video is cartoonish and child-friendly, with bright colors and simple shapes."
+    ],
+    "video_ids": [
+        "LpkylclhRpU_91_0to122"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A lion, a dinosaur, an elephant, and a duck (stuffed animals).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features four stuffed animals: a lion (brown with a mane), a green dinosaur (reptilian appearance), an elephant (gray with large ears and trunk), and a yellow duck. These match the specified objects in the condition. The presence of a pink pig is an additional element, but it does not contradict the core description, as the condition only requires the four specified animals to be present, which they are.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows four distinct animals: a lion,orange with a mane), a dinosaur dinosaur,possiblysemblingilian),), an elephant (gray with large ears), trunk), and a duck duck ( These animals the ' objects, the '. The setting of musical blue andlet not additional element but but it does not contradict the core description of as the question only requires the presence specific animals.\"\n be present.\"\n not they are.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..19fce05368fbe383cee89760178f255694fb441b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c4e9923575114ea151c185a9e48894afef4eb5651a4f24f9300d8370ab63436c
+size 132283
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..eeaced270bdf7b1f930c276e319457ba326937da
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the growth of a tomato plant in a series of three frames. In the first frame, the plant is just beginning to sprout, with a few small green leaves peeking out from the soil. In the second frame, the plant has grown significantly, with several green tomatoes hanging from the vine. The third frame shows the plant in full bloom, with a mix of green and yellow tomatoes hanging from the vine, indicating that they are ripe and ready for harvest. The plant is housed in a white plastic container, and the entire scene is set against a backdrop of a concrete floor. The video is a simple yet effective representation of the growth and development of a tomato plant."
+    ],
+    "video_ids": [
+        "rgXDZ9zcUu8_12_0to128"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A tomato plant? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a tomato plant with multiple tomatoes at various stages of ripeness, growing in containers. The plant's stems, leaves, and fruit are visible, and the overall scene is consistent with a tomato plant in a garden or greenhouse setting. There are no elements that contradict the description of a tomato plant.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a tomato plant with green green in various stages of ripeness, growing in a. The presence is structure, leaves, and tomatoes are visible, which the setting scene is consistent with the tomato plant. a domestic or indoor setting. The are no elements in contradict the description of a tomato plant.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cab04d6e98e0ff97e2f42213908f5bc5d09dc130
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:970876b776ccfdb2ae797cbe907f6328e580fae4b4b2e5cd28b5c764fa0d58ea
+size 329641
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..e29fd2acb15cf2ed3201a0688862fac38684af36
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the interior of a Ferrari car, showcasing its luxurious design and advanced technology. The dashboard is adorned with a variety of gauges and screens, including a central infotainment system displaying the Ferrari logo. The steering wheel is equipped with multiple buttons and dials, indicating the car's high-performance capabilities. The interior is upholstered in a rich, brown leather, and the overall design exudes a sense of luxury and sophistication. The video is likely a promotional or informational piece, highlighting the features and design of the Ferrari car."
+    ],
+    "video_ids": [
+        "UtR_aizjJE0_10_22to201"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dashboard, central infotainment system, Ferrari logo, steering wheel, buttons, dials, leather upholstery? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows the dashboard, central infotainment system, Ferrari logo (on the steering wheel and display), steering wheel, buttons, dials, and leather upholstery. All specified elements are clearly visible and accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a dashboard, central infotainment system, Ferrari logo,on the steering wheel), center), steering wheel, buttons, dials, and leather upholstery. The these elements are present visible and match represented in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a430e44fedbf96b19e0e1b282d950fe28b55ef9e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f2789b2555594add89fb1c2ab34d3345553b60f73c448d334d3f9232a823f7a7
+size 196024
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..351d4e657a23512ff09d724cdbc582352a06a2c0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a lively scene at a Chinese temple. A man, dressed in a black jacket and carrying a helmet, stands in front of the temple, seemingly in awe of the grandeur of the structure. The temple, with its yellow roof and intricate designs, is a striking feature in the image.   The temple is surrounded by a bustling crowd of people, some of whom are carrying umbrellas, adding a splash of color to the scene. The crowd is a mix of individuals, some walking around, others standing still, all seemingly engrossed in the beauty of the temple.  The temple is nestled amidst lush green trees, creating a serene backdrop to the vibrant activity in the foreground. The overall scene is a beautiful blend of tradition and modernity, with the ancient temple standing tall amidst the hustle and bustle of the crowd."
+    ],
+    "video_ids": [
+        "X4wsDhTkB9A_8_0to134"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black jacket and helmet, a temple, a crowd of people? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a man in a black jacket holding a helmet, a temple-like structure with Chinese characters, and a crowd of people in the background. All core elements described are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a man in a black jacket and a pink, a temple with structure with a architectural, and a crowd of people holding the background. The the elements of in present, match depicted in}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1929a011dcaef568ff7b2ad6952291205a06376d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:def03b5d270f7a762f02f3df659a5cd9f13515237e0ef8cabb8af9f969908f03
+size 215582
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..ae79eb5bebea6ee23dfa369c8b99efaf1d7575fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a black and brown dachshund dog sitting on a grassy lawn. The dog is wearing a blue and white shirt. The dog's eyes are wide open, and it appears to be looking directly at the camera. The background is blurred, but it seems to be a natural outdoor setting with grass and trees. The dog's expression is calm and attentive. The style of the video is a close-up shot of the dog, focusing on its face and upper body. The lighting is natural, suggesting that the video was taken during the day. The overall mood of the video is peaceful and serene."
+    ],
+    "video_ids": [
+        "U_jaoMlyk68_23_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black and brown dachshund dog? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a black and brown dachshund dog wearing a harness, sitting on grass with trees in the background. The dog's appearance, including its coloration and breed characteristics, matches the description. The text overlays do not contradict the object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a close and brown dachshund dog, a blue, which on grass. a in the background. The dog's color and including its color and and breed,, align the description of The presenceural and not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1a52baa1040185c1df330c395c537997c1d91602
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c8d8f3ef354f4b24e4954d3980c534ca9a73e460b8f8fc6a325625bb358ca6e2
+size 169632
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..837324cddd713ffba24152ec4d964b02b39edb0a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features two men standing outdoors near a body of water, possibly a lake or river, with a forested background. Both men are wearing sunglasses and casual clothing, with one wearing a cap. They appear to be in a relaxed, outdoor setting, possibly enjoying a leisurely day. The style of the video is casual and informal, capturing a moment of camaraderie between the two men. The focus is on the subjects, with the natural environment serving as a backdrop. The lighting suggests it might be daytime, and the overall mood of the video is laid-back and friendly."
+    ],
+    "video_ids": [
+        "-Cgl2_yxUcY_0_0to101"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men wearing sunglasses and casual clothing, one wearing a cap.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men wearing sunglasses and casual clothing. One of them is wearing a cap, and both are dressed in relaxed attire (a t-shirt and a button-up shirt). The background and their attire do not contradict the description, and the core elements are clearly present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men standing casual and casual clothing, One of the is wearing a cap, which they are standing in a attire suitablet blue-shirt and shorts polo-up shirt with The setting includes setting poses align not contradict the description, and the presence elements of present visible.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..14ea82f51f7fd29876b5563ae40ffe6254fed584
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e41471dfaa08d6de77580af9823dd18edf12ca4df654c9bed80bfefb199fcf3
+size 323773
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..37685a3d528375c814e9d3e5d5c00fa6fffd7f35
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a chef in a professional kitchen, preparing a dish. The chef is wearing a white chef's coat and glasses, and is focused on the task at hand. The kitchen is well-equipped with stainless steel appliances, including a stove and an oven. There are also several pots and pans hanging from a rack above the stove. The chef is using a wooden spoon to mix ingredients in a bowl. The overall style of the video is realistic and informative, likely aimed at teaching viewers how to prepare a specific dish. The chef's actions are the main focus of the video, with the kitchen serving as a backdrop."
+    ],
+    "video_ids": [
+        "eOV9CmpA69o_30_0to120"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A chef in a white chef's coat and glasses, a wooden spoon, a bowl, pots and pans hanging from a rack.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chef wearing a white chef's coat and glasses, using a wooden spoon to mix ingredients in a bowl. Pots and pans are clearly visible hanging from a rack in the background. All core elements described are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a chef in a white chef's coat and glasses, stirring a wooden spoon to stir ingredients in a bowl. Theots and pans are hanging hanging hanging from a rack in the background, The elements elements of in present and accurately depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..60b15f6c43097733cabdd7889b1f9ad540e26b47
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1363-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67e98856772e6ac05b2443da7ef94a355197ef6fbc006ff17bafd740b51dcff1
+size 165671
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..0eda6dd07da3a5489a20f48371f1f660bfd3bb33
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a group of police officers in a city street. The officers are dressed in blue uniforms with the word \"HANDHAVING\" printed on the back, indicating they are from the Netherlands. They are standing near parked cars and a brick building. The officers appear to be engaged in a conversation, possibly discussing a situation or providing directions. The street is lined with parked cars and the buildings have a traditional European architecture. The overall style of the video is a real-life, candid capture of a moment in the officers' day."
+    ],
+    "video_ids": [
+        "Avtpf43mkm4_8_20to170"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A group of police officers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a group of police officers in uniform, identifiable by their blue jackets with 'HANDHAVING' written on the back and police insignia. They are interacting with each other in an urban setting, which aligns with the description of a group of police officers.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows depicts two group of two officers, uniform standing standing by their blue shirts and 'POLLINGN' written on the back. their capsia. They are standing with each other in a urban setting, which aligns with the description of a group of police officers.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..367a6d63c18a25a0aadd193ef72d2299948128d5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f1b0713c1cf6045d29b96ab6a222a1928da9469d3204a0c8f5414f8ad5ffec7
+size 236537
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..d54e4fce66edc4890827e323ace0e5e39ba95cf4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a gray mouse with pink ears is seen enjoying a snack. The mouse is holding a green cup in its mouth and a piece of cake in its hand. The mouse is standing on a pink chair, which is placed on a yellow surface. The background features a green wall with a blue plate on it. The mouse appears to be in a happy and relaxed state, enjoying its food. The video captures a cute and playful moment of the mouse."
+    ],
+    "video_ids": [
+        "8y6RXCCZd1o_118_60to192"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A gray mouse with pink ears, a green cup, and a piece of cake.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a gray mouse with pink ears, holding a green cup and a piece of cake. These core elements match the description exactly, and no contradictory elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a gray mouse with pink ears, holding a green cup. eating piece of cake. The elements elements match the description provided. and there additional elements are present.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9d2868ccf984b509af8486a09f36c465365ed88b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a28e65c9c985581053d20d5069e3bb715d6a766141e7ee192bf778e66bca39d3
+size 115387
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..7b896dacfe17abcbdacce8e3110876e051e8723f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene outdoor scene set in a lush garden. The focal point is an ice sculpture of a bird, intricately carved to resemble a dove or pigeon, lying on its side amidst scattered ice fragments. The sculpture is positioned on a patch of vibrant green grass, surrounded by various plants and shrubs that add depth and texture to the background. The camera remains stationary throughout the sequence, allowing viewers to take in the tranquil setting and the delicate details of the ice sculpture. The lighting suggests it is daytime, with sunlight casting soft shadows across the scene, enhancing the natural beauty of the environment. There are no visible characters or significant changes in the scene, maintaining a peaceful and contemplative atmosphere."
+    ],
+    "video_ids": [
+        "261bdd222ee3cfc1967f4e1954ac9b892088c904a11625d605be399f5191471e"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An ice sculpture of a bird (dove/pigeon) and scattered ice fragments, vibrant green grass, various plants and shrubs.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an ice sculpture that resembles a bird (likely a dove or pigeon) lying on vibrant green grass, surrounded by scattered ice fragments. The background includes various plants and shrubs, matching the description. There are no conflicting elements in the scene that contradict the specified conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts an ice sculpture of resembles a bird,likely a dove or pigeon) and on a green grass. which by various ice fragments. The background includes various plants and shrubs, which the description provided The are no additional elements in the video.\"\n would the given conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4e4da728e2d8d6e06b9da089f253b304c3b8d543
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e3d1edde61e11761a0a290a8bf73d4a435f36537a676774f049d9ed298dc5b47
+size 155470
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..debf2f5ba5bcdf610530b400e502546ad1250f53
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of an orange SUV parked on the side of a road. The vehicle is facing the camera, and its license plate is clearly visible. The SUV has a distinctive front grille and headlights, and it is equipped with large, black wheels. The background of the video features a grassy area and a tree, suggesting that the SUV is parked in a suburban or rural setting. The style of the video is straightforward and documentary-like, focusing on the vehicle without any additional context or embellishments."
+    ],
+    "video_ids": [
+        "vr-KYsPh-k4_42_0to186"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An orange SUV with a distinctive front grille, headlights, and large, black wheels. Its license plate is clearly visible.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an orange SUV with a distinctive front grille, sleek headlights, and large, black wheels. The license plate is clearly visible and readable as 'JK9 72F'. All elements described in the condition are present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an orange SUV with a distinctive front grille, headlights headlights, and large black black wheels. The license plate is clearly visible, positioned, well1 9172'. The these of in the condition are present and match depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b344340320899498caa4dafc06570dca589b44d1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da2a8c48ed6b9757c208965e4a2dc2cb3128ac4946b39ad35a3fc055f52c04a4
+size 211704
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..33905ca4f9845e82495467faf6bc200c7d30f514
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are seen standing together in a garden, smiling and enjoying each other's company. The garden is lush with green plants and trees, providing a serene and peaceful backdrop. The man is wearing a purple shirt, while the woman is dressed in a pink top. Their expressions are joyful and content, suggesting a strong bond between them. The video captures a moment of happiness and tranquility in a beautiful outdoor setting."
+    ],
+    "video_ids": [
+        "Rr4zgX_2Sz0_58_0to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman standing together, smiling and enjoying each other's company.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a woman standing together in an outdoor garden setting. They are smiling and appear to be enjoying each other's company, as evidenced by their facial expressions and body language. The background elements (trees, foliage) do not contradict the core description and are consistent with a natural outdoor environment.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a woman standing close in a outdoor setting setting. Both are both and appear to be enjoying each other's company, which indicated by their body expressions and body language. The man is,green and green) do not contradict the description description and are consistent with the garden, environment where}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..38131b4dddf9f0d5d880c02058acf89afb167266
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cecb2874b1276c83bbb7407567c57ad462f98dacdaf6cac53e5d3c0eaa851a87
+size 173197
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..d4043d3a19f71f00ceed66bb10c9562b98859a56
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a suit and tie sitting in the driver's seat of a car. He appears to be in a showroom or exhibition, as evidenced by the bright lights and other cars in the background. The man is adjusting his tie, suggesting that he is preparing for a presentation or demonstration. The car's interior is well-lit, highlighting the sleek design and modern features. The man's attire and the setting suggest a professional or business context. The video captures a moment of preparation and anticipation, as the man prepares to showcase the car to potential buyers or attendees."
+    ],
+    "video_ids": [
+        "3aLwB-Uhi4Q_6_30to178"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, a car (driver's seat).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man dressed in a suit and tie sitting in the driver's seat of a car. The setting appears to be an auto show, with other cars and people visible in the background, but the core elements \u2014 the man in a suit and tie and the car's driver's seat \u2014 are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man in in a suit and tie, in the driver's seat of a car. The man and to be a indoor show or as the cars visible people visible in the background, but the core elements of a man in a suit and tie, the car's driver's seat \u2014 are present depicted.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..87a3ac7ad1848b48833c284513a49ff8543d1b3e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a0166885166f7dc0d32ba93e5d770b0f76b5b0c3df64f20fde016a7ebd86a14
+size 153038
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..7fec6b6567093fa18ff14e686bc7546b2967a12c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a split-screen interview with a football player and a commentator. The football player, wearing a red and black jersey with the logo \"UEFA\", is shown celebrating with his arms outstretched, a wide smile on his face. The commentator, wearing a black shirt and glasses, is seen with a concerned expression, looking off to the side. The background is a green network of lines and dots, suggesting a digital or analytical theme. The video is likely a sports commentary or analysis segment, focusing on the player's performance or a recent match."
+    ],
+    "video_ids": [
+        "h6jgURLPZ9A_17_32to252"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A football player and a commentator? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a football player wearing a Belgium national team jersey and a commentator in a black shirt, both of whom are clearly visible and central to the scene. The background and other elements do not contradict the core description of having a football player and a commentator.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a football player wearing a red national team jersey with a commentator wearing a black shirt. which of which are positioned visible and fulfill to the scene. The presence is lighting elements, not contradict the core description of a a football player and a commentator.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..581f83aa717a31bd2b647ecc3270979022514c4c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3657fc6db84876f53081f04842fe6a44ec238757c5d6bdb17f53b1bf85ab0acb
+size 191329
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..e39bdfa7ae4ab225d6ab84913c64bb710fa531f4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man wearing glasses and a blue shirt is seen interacting with a small dog. The dog, with its white and brown fur, is sitting on a pink surface. The man appears to be speaking to the dog, possibly giving it a command or a treat. The scene is set against a white wall, which provides a neutral backdrop for the interaction between the man and the dog. The overall style of the video is casual and intimate, capturing a moment of connection between the man and his pet."
+    ],
+    "video_ids": [
+        "FMF7hHlb-MQ_35_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man wearing glasses and a blue shirt, a small dog with white and brown fur? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing glasses and a blue shirt, and a small dog with white and brown fur in the foreground, even though the dog is out of focus. These core elements match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man wearing glasses and a blue shirt, and a small dog with white and brown fur. a foreground. which though the dog is not of focus. The elements elements match the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ff7ea3339a4ff813732a2f70aba666f73920c6c8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1364-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95f63b61942f256a7a215e65db289dffac1d0f52ce3613b55f4d18cc6e506f44
+size 138279
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..6278cfc794b3eeb4cccbd30916d336406cc80d87
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a German Shepherd dog standing on a beach. The dog is wet, suggesting it has been in the water. The dog's ears are perked up, indicating alertness. The background shows a body of water, likely the ocean, with a clear sky above. The overall style of the video is naturalistic, capturing the dog in its natural environment. The focus is on the dog, with the background serving as a complementary element to the scene. The video does not contain any text or additional objects. The dog's position and the direction of its gaze suggest it is looking towards the camera. The video does not contain any action or movement, but rather captures a still moment in the dog's day."
+    ],
+    "video_ids": [
+        "aaia0tiJYOA_45_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: German Shepherd dog? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a German Shepherd dog, identifiable by its distinct features such as pointed ears, black and tan fur, and facial structure. The dog remains the central focus throughout the video, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a German Shepherd dog standing which by its distinctive black such as the ears, a and tan fur, and the markings. The dog is the central focus throughout the frames, and the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1c57b4172612cc173132f192b11c88361d62d8b9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ddf327e7ea2b0fba145d49dd8109161ec1de412594498dac948d3956a40968a9
+size 212140
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..8b423bbd0ca88192e48c85706352172c690233e1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with blonde hair, wearing a gold necklace and a beaded dress, sitting at a table with a menu in front of her. She is holding a glass of water and appears to be speaking. The setting is a restaurant with a blue and white color scheme, and there are wine glasses and a bottle on the table. The style of the video is a candid, real-life scenario, possibly from a television show or a documentary."
+    ],
+    "video_ids": [
+        "FU1NKB3Ge20_0_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with blonde hair, wearing a gold necklace and a beaded dress, holding a glass of water and sitting at a table with a menu in front of her. She appears to be speaking.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair, wearing a gold necklace and a beaded dress, sitting at a table with a menu in front of her. She appears to be speaking, and there is a glass of water on the table. Additional elements like a trophy and a person holding a box are present but do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a woman with blonde hair, wearing a gold necklace and a beaded dress, holding at a table. a menu in front of her. She is to be speaking and holding she is a glass of water and the table. The elements like a wine and a wine in a glass are present but do not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c2db15334c591e45d1eea454a8766b322bd62d0a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:834151c4dcc65c7393a09678956a096aa0c20fae4c85afc94842bc7a2f3c0181
+size 212449
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..a09a636874dbb088266a0a9353d392cafdaf9489
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young man is seen in a living room, holding a black object in his hands. He is wearing a black hoodie with a green logo on it and a black baseball cap. The living room is well-lit with natural light coming in through the windows. The man appears to be in motion, possibly dancing or playing with the object. The overall style of the video is casual and fun, capturing a moment of joy and playfulness."
+    ],
+    "video_ids": [
+        "TiNturbjXBg_4_254to394"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man, a black object in his hands, a black hoodie with a green logo, a black baseball cap.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young man wearing a black hoodie with a green logo and a black baseball cap, holding a black object (likely a phone or remote) in his hands. All core elements described are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man wearing a black hoodie with a green logo and a black baseball cap. holding a black object inlikely a gaming or a) in his hands. The the elements of in present and match with the video content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d1421ca1a4a3774dc3ec45dc52589e548d776e21
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d0e794bcd3fcf3c0301f5ec7730773a2657cb3f6c6384e96bd740d228551ae87
+size 182046
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..a16b60e9a1e9f8946c6b39853b0fc27a7d6cdc44
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video begins with a close-up shot of a dark-colored beetle resting on a sandy surface. The beetle is positioned centrally in the frame, with its body oriented diagonally across the sandy ground. Its legs are spread out, and it appears to be stationary. The sand around the beetle is fine-grained and light brown, indicating a dry environment. The beetle's shell is shiny and reflective, suggesting it might be wet or coated with a glossy substance. The background is a uniform sandy texture, emphasizing the isolation of the beetle.  As the video progresses, there is no significant movement from the beetle itself. However, the scene transitions to show the same sandy area, but now the beetle is no longer visible. Instead, a small, reddish-brown insect, possibly an ant, is seen moving across the sand. This insect is much smaller than the beetle and has a more elongated body shape. It appears to be crawling quickly, leaving a trail of disturbed sand behind it. The camera remains"
+    ],
+    "video_ids": [
+        "f39a1c4a0f8591b5e84428f7d872565efe313a7769e5e15d422fc01d0ecde776"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A dark-colored beetle and a small, reddish-brown insect (possibly an ant).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dark-colored beetle on sand, and a small, reddish-brown insect (likely an ant) briefly interacts with it. The core objects described are present, and no major contradictions exist.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dark-colored beetle and a, which a small, reddish-brown insect,possibly an ant) moving appears with the. The presence elements described in present, and the additional contradictions are.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3f970d45c538bc4898894ef08c9a12ee6d5dba90
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0d6a4402fb68da5b0b7066fb0b9d80a384f4f52247c18e056691ada646fe4d2a
+size 239957
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..e4a3b6a4f5ff7af48397d66dbf68d9298e91ebf2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the majestic beauty of a lighthouse perched on a rocky cliff. The lighthouse, standing tall and white, is the focal point of the scene. It's surrounded by a rugged coastline, where the waves crash against the rocks, creating a dynamic and powerful display of nature. The sky above is a clear blue, dotted with fluffy white clouds, adding a serene backdrop to the scene. The lighthouse, with its beacon, stands as a symbol of guidance and safety amidst the raw power of the ocean. The video is a testament to the awe-inspiring beauty of nature and the human-made structures that coexist with it."
+    ],
+    "video_ids": [
+        "IwTnKz7Z0mk_28_0to166"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A lighthouse, waves crashing against rocks, and a beacon.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a lighthouse situated on a rocky cliff overlooking the ocean, with waves visibly crashing against the rocks. The lighthouse itself functions as a beacon, fulfilling the core elements described in the condition. Additional elements such as nearby buildings and trees do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a lighthouse situated on a rocky out, the ocean. with waves crashing crashing against the rocks. The presenceighthouse is serves as a beacon, which the ' elements of in the question. The elements such as the vegetation and a are not contradict the main and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1d084024743a10e1bfdb59c26316a50d31071612
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83431804d8758126d4bb9fe839199437177e6106a324890b3b80af81880b4dff
+size 144039
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..7a5e22ed93cbd004d44dc33bcbdbcc017e9ae3f7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a silver Porsche sports car parked in a garage. The car is positioned in the center of the frame, with its rear facing the camera. The car's design is sleek and modern, with a large rear wing and a distinctive taillight design. The car is parked next to a yellow line on the floor, indicating a designated parking spot. The garage itself is spacious and well-lit, with a high ceiling and a concrete floor. The car's license plate reads \"MR03 JN\". The style of the video is straightforward and documentary, with no additional elements or effects. The focus is solely on the car and its surroundings."
+    ],
+    "video_ids": [
+        "DJMBapnPnLg_10_0to113"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A silver Porsche sports car and a yellow line on the floor.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a silver Porsche sports car parked indoors, with a yellow line visible on the floor beneath it. The car's model is identifiable as a Porsche 911 GT3, and the yellow line is prominently displayed, matching the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a silver Porsche sports car, in, and a yellow line on on the floor. it. The car's design and identifiable by a Porsche,911,3, and the yellow line is a displayed, matching the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1ea5e614cae687fee81d954f37f96bf5bc0efcc3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b8233c989d12652f7fd143ca97d18dbe6cfe8a7e31266d3b364fa5ac14b7c75
+size 103708
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..bba9b18a9cb479a26c95db72be975c81ce37e5bc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man enjoying a meal at a dining table. He is wearing a red and black hoodie with the word \"OUTCAST\" on it. The man is seated at a dining table with a bowl of soup in front of him. He is using a spoon to eat the soup. The table is set with various items including a cup, a fork, and a knife. The setting appears to be a restaurant or a dining room. The man seems to be enjoying his meal and is smiling while eating. The overall style of the video is casual and relaxed, capturing a simple yet enjoyable moment of the man's day."
+    ],
+    "video_ids": [
+        "CfcUwQB2ndI_147_202to372"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a dining table, a bowl of soup, a spoon, a cup, a fork, and a knife.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man seated at a dining table, eating from a bowl of soup with a spoon. A cup, fork, and knife are also visible on the table, fulfilling all the specified object conditions. The scene is consistent with the description without any contradictory elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man sitting at a dining table. eating from a bowl of soup using a spoon. There cup is a, and knife are also visible on the table. fulfilling the the conditions objects conditions. The presence is consistent with the description provided any contradictions elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d52c73242163418f3d1e4abf440ab5456a4961d0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a7d80126814a16d4c53041f2b6cc86ba14a57414d0a49f2bad1094734628ba9c
+size 143295
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..c67d30cb4dddf8c38ec2b86067f4c1e1e9f910c4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a professional football player in action. He is dressed in a vibrant red uniform with the number 15 prominently displayed. The player is seen holding a football in his right hand, ready to throw it. His stance suggests he is in the middle of a game, possibly preparing to make a crucial pass. The background is a blur of spectators, indicating that the game is being played in a stadium filled with fans. The focus is solely on the player, emphasizing the intensity of the moment. The style of the video is dynamic and action-packed, capturing the essence of the sport."
+    ],
+    "video_ids": [
+        "_Ztkg8IJskQ_85_24to160"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A professional football player in a vibrant red uniform with the number 15, holding a football in his right hand.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a professional football player wearing a vibrant red uniform with the number 15, holding a football in his right hand. The player is identifiable as Patrick Mahomes of the Kansas City Chiefs, and the uniform, number, and football are accurately represented. The background is blurred, but it does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a professional football player wearing a vibrant red uniform with the number 15. holding a football in his right hand. The player is also as a Mahomes, the Kansas City Chiefs, as the scene and number, and action are clearly depicted. The player, a, which it does not contradict the description description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..396852184f8cc7489d50a65407d7f8feb52e74f2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1365-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f0cb34a15dedec5b1fe4cba8debf37e899ea3365a5a55d8e62541cc17105e79c
+size 214808
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..c61d77abdfc826bf1f8a8fb24cba923fade4b3b2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the breathtaking view of a rocky coastline under a clear blue sky. The camera pans across the scene, showcasing the towering cliffs that rise majestically from the deep blue water. The cliffs, with their rugged texture and jagged edges, are a testament to the power of nature. The water, a deep shade of blue, is calm and serene, reflecting the clear sky above. The video is a beautiful representation of the natural beauty of the coastline, capturing the essence of the location in a single frame."
+    ],
+    "video_ids": [
+        "TGkmMusQL9g_24_0to143"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Cliffs, water, sky? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features cliffs rising from the water, a vast expanse of water in the foreground, and a clear blue sky with scattered clouds. These elements fully satisfy the 'Object(s)' condition as described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts features cliffs, steep the water, with body expanse of water, the middle, and a clear blue sky in no clouds. These elements align align the 'Object(s)' condition as described.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..76260f00edef3a0eab804e9a054aa6e54b6a3b09
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9695084f984864ab5bdf3978fe101811ced374265c7c32e98daacb163f7952f7
+size 152901
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..27610e54363ca2fad9be865e3cb9820cc690fd34
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two young women are seated at a table in a library, engrossed in their work. The library is filled with books, and the women are surrounded by a quiet atmosphere. The woman on the left is wearing a black leather jacket and glasses, while the woman on the right is dressed in a yellow blouse. They are both focused on their notebooks, which are open in front of them. The scene captures a moment of concentration and academic diligence."
+    ],
+    "video_ids": [
+        "7lNYfZnp758_4_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two young women, a table, notebooks, books.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two young women sitting at a table with notebooks, which aligns with the 'Object(s)' condition. Although there are additional elements like a man walking by and bookshelves in the background, these do not contradict the core description and are consistent with a library setting. The presence of books on the shelves further supports the context.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two young women sitting at a table, notebooks and each ares with the 'Object(s)' condition. The there are additional books like books bookne in in ashelves in the background, these do not contradict the core description and are acceptable with a library or. The presence of books on the table further supports the library of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..79c71487cf8069f9dac7f94b410e9e57a79f4096
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3594bbb9e8221d86caf184892c3790a209f5dfc7f9c7903d0b9f77a1fa7fd30
+size 100514
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..beb46769e1ef05d3a30a25f80fe4bedaa2c6f73e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a bald man with a beard and glasses, sitting on a purple couch in a room with purple lights. He is wearing a black shirt and a necklace. The room has a bookshelf filled with books and various objects, and a guitar is visible in the background. The man appears to be speaking or singing, as he is holding a microphone close to his mouth. The overall style of the video is casual and intimate, with a focus on the man and his surroundings."
+    ],
+    "video_ids": [
+        "8aCGL9GpVUg_3_147to346"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bald man with a beard and glasses, wearing a black shirt and a necklace, holding a microphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man with a beard and glasses, wearing a black shirt and a necklace, sitting on a purple couch. He appears to be speaking or presenting, and although a microphone is not visibly held in his hands, a small lavalier microphone is clipped to his shirt, which aligns with the description of him holding a microphone. The core description is largely fulfilled.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man with a beard and glasses, wearing a black shirt and a necklace. holding in a couch couch. He is to be speaking or singing, as at he microphone is not directly held in his hand, the microphone microphoneavalier microphone is attached to his shirt, which iss with the description of holding holding a microphone. The background elements is largely matched.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cb5a2215c2873bbe14ded5eb08acafee247572a3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4549f2484588297daf8dfed43ff398a6650d25fa8100f71d0c29c97e33ad471
+size 182278
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..aa7d59cf449786f74ffe9b74f48610dc0c0c0817
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene outdoor scene set in a lush garden. The focal point is an ice sculpture of a bird, intricately carved to resemble a dove or pigeon, lying on its side amidst scattered ice fragments. The sculpture is positioned on a patch of vibrant green grass, surrounded by various plants and shrubs that add depth and texture to the background. The camera remains stationary throughout the sequence, allowing viewers to take in the tranquil setting and the delicate details of the ice sculpture. The lighting suggests it is daytime, with sunlight casting soft shadows across the scene, enhancing the natural beauty of the environment. There are no visible characters or significant changes in the scene, maintaining a peaceful and contemplative atmosphere."
+    ],
+    "video_ids": [
+        "261bdd222ee3cfc1967f4e1954ac9b892088c904a11625d605be399f5191471e"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An ice sculpture of a bird (dove/pigeon) and scattered ice fragments, vibrant green grass, various plants and shrubs.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows an ice sculpture that resembles a bird (likely a dove or pigeon) lying on vibrant green grass, surrounded by scattered ice fragments. The background includes various plants and shrubs, matching the description. There are no conflicting elements in the scene that contradict the specified conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a ice sculpture that resembles a bird,possibly a dove or pigeon) with on vibrant green grass. surrounded by various ice fragments. There presence includes various plants and shrubs with which the description. The are no additional elements in the video.\"\n would the given conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2a3d85f29e2dcdd814df9e45ff8e5c065eb2e1c4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d88c7fcbcd73ad84a47a51639a4fe2ca41f03deb8316c4872b4cc8376396563e
+size 152759
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..dbc5ee544735b5737ca23b2fa4238f088c0fb16d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman and a young girl are in a kitchen, preparing food together. The woman is standing at a counter, stirring a bowl of food, while the girl is sitting on a stool, holding a teddy bear. The kitchen is well-equipped with a refrigerator, an oven, and a sink. There are also potted plants and a vase on the counter. The woman is wearing a blue sweater, and the girl is wearing a pink shirt. The overall atmosphere of the video is warm and inviting, capturing a moment of bonding between the woman and the girl as they cook together."
+    ],
+    "video_ids": [
+        "Hr96wp-ZxaQ_31_0to156"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a young girl, a bowl of food, a teddy bear, a counter, a stool, a refrigerator, an oven, a sink, potted plants, and a vase.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a woman, a young girl holding a teddy bear, a bowl of food on a counter, a refrigerator, an oven (visible in the background), potted plants, and a vase (on the counter). Although a stool is not visibly present, all other specified objects are present and consistent with the description. The presence of additional elements does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a woman, a young girl, a teddy bear, a bowl of food, the counter, a stool, a oven,part in the background), aotted plants, and a stool.on the counter). The there sink and not visible present, the other objects objects are present and match with the description.\"\n The presence of additional elements, not conflict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d076f5a01107a0962582132f8a0a573138e74192
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dbdf4b555bc06e6edf0015c9502bcecbf080121961e8c36515c801f5874e9883
+size 123795
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..c1106348bfee8721fda41f4deb7e5fcdd0aa5785
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man with a beard and a shaved head, wearing a gray sweatshirt with the word \"Vegan\" printed on it. He is standing on a sidewalk, holding a red apple in his hand, and appears to be taking a bite from it. The man is wearing earbuds and has a relaxed posture. In the background, there are parked cars and a clear blue sky. The style of the video is casual and candid, capturing a simple, everyday moment."
+    ],
+    "video_ids": [
+        "0Zd8b2aEe1M_8_22to145"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man with a beard and shaved head, wearing a gray sweatshirt with 'Vegan' printed on it, holding a red apple.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and shaved head, wearing a gray sweatshirt with the word 'VEGAN' printed on it. He is holding a red apple, which is partially eaten, and appears to be taking a bite. The core description is accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and a head, wearing a gray sweatshirt with ' word 'VGAN' printed on it. He is holding a red apple and which align consistent eaten. and appears to be in a bite. The setting elements of largely represented in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c503e6fe1979e47aa684f5991c564b3149442bc4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:09fa97fc6488ae6edb48548b9b47fda5642eff1c986a3c2e455c328f7f8fb75e
+size 223758
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..eac833c9ac42abfccde110c8ccdc05bff732c66a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a young boy with blonde hair, wearing a red t-shirt with the words \"Old Navy\" and an American flag design on it. He is standing in a kitchen, with a stainless steel refrigerator and a white cabinet visible in the background. The boy is looking down at the camera, and his expression changes from a neutral one to a slight frown. The style of the video is casual and candid, capturing a moment in the boy's life. The focus is on the boy and his expression, with the kitchen setting providing context."
+    ],
+    "video_ids": [
+        "CWYNU2TPR08_19_19to167"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young boy with blonde hair wearing a red t-shirt with 'Old Navy' and an American flag design.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young boy with blonde hair wearing a red t-shirt that clearly displays the 'Old Navy' logo and an American flag design. The description is accurately reflected in the visual content of the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young boy with blonde hair wearing a red t-shirt that features displays the 'Old Navy' logo and an American flag design. The description matches largely represented in the video content of the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9bbc2bb9775eca8f7a77cb7ec619a77d7f122d82
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b5d2143fed1aacd4f58b06092db6237a1fcde755f55ccc969b95351a875bdc4
+size 247016
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..fecf4b4cb218c8ef9162cb57a264b5f781d91d1a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a chocolate dessert being cut with a knife. The dessert is placed on a white plate, and the knife is being used to slice through the chocolate. The chocolate has a glossy finish and is adorned with chocolate chips. The knife is silver and has a pointed tip. The background is blurred, but it appears to be a kitchen setting. The style of the video is a slow-motion shot, capturing the details of the dessert and the knife as it slices through the chocolate."
+    ],
+    "video_ids": [
+        "Iv4mQJsBTPk_331_775to958"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chocolate dessert, knife, chocolate chips? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chocolate dessert in the form of a high-heeled shoe, which matches the 'Chocolate dessert' condition. There are no knives visible in the video, but since the presence of additional elements is acceptable as long as they don't conflict with the core description, this is not a contradiction. Chocolate chips are not explicitly visible, but the small chocolate droplets being dispensed can be considered a form of chocolate chips or similar chocolate elements. Thus, the core description is largely fulfilled.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close dessert being the form of a dome domeeled shoe, which is the 'Chocolate dessert' condition. A is chocolate additional or in the video, but the the question of a elements is acceptable, long as they do't conflict with the core description, the can not a contradiction. The chips are not present mentioned, but the dessert pieces piecesplets on cutensed from be interpreted as form of chocolate chips, chocolate decorative decorations, Therefore, the video elements is largely matched.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..57d552ba9dc3c7bdc53e430874cecc629e8d965e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1366-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75bf25ce292aacd792d8f655c10345c81688373bf4164fb1af2d2c85ae00acf0
+size 146030
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..e7f4297dc8163e2ee3ed2bd2910e47568b3df137
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a jubilant moment on a football field. A football player, wearing a white jersey with the number 10, is the central figure. He is smiling broadly, his tongue sticking out in a playful manner, as he walks off the field. His joy is palpable, suggesting a recent victory or a successful play.  In the background, a crowd of spectators can be seen, their faces a blur of excitement and anticipation. The field itself is a vibrant green, contrasting with the player's white jersey. The atmosphere is one of celebration and camaraderie.  The style of the video is dynamic and energetic, capturing the essence of the sport. The focus is on the player, with the crowd serving as a backdrop to his triumphant moment. The video does not contain any text or additional elements, allowing the viewer to fully immerse themselves in the scene. The overall impression is one of a thrilling sports event, filled with action and emotion."
+    ],
+    "video_ids": [
+        "ONs2OPfG2GM_12_27to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A football player in a white jersey with the number 10, a crowd of spectators in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a football player wearing a white jersey with the number 10, and there is a crowd of spectators visible in the background. These elements match the core description provided. Additional elements such as cameras and reporters do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a football player wearing a white jersey with the number 10, which there is a crowd of spectators in in the background. The elements match the description description provided, The elements, as the and the in not contradict the main and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1374693edeb727ceed645c2fa3eac4f3e244f257
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:043806d04d6eabb67bd1f6c4cb55a6ae1549d5f980a13aeb6b8642eadf67448d
+size 217951
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..e9dcbbfaa441b3ed34676cf2ce8f8c4bf8d849b9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a two-story house with a white exterior and a gray roof. The house has a front porch with a railing and a door. There are two windows on the second floor and two on the first floor. The house is surrounded by a well-maintained lawn with a sidewalk leading to the front door. A for sale sign is placed in the front yard, indicating that the house is on the market. The sign has a phone number and the name of the real estate company. The style of the video is a real estate listing, showcasing the house and its surroundings to potential buyers. The video is likely intended to highlight the house's features and appeal to potential buyers."
+    ],
+    "video_ids": [
+        "Xrn1GBSpbZI_1_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two-story house, front porch with railing and door, two windows on the second floor, two windows on the first floor, well-maintained lawn, sidewalk, for sale sign? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a two-story house with a front porch, railing, and door. There are two windows on the second floor and two windows on the first floor. The lawn is well-maintained, a sidewalk is visible, and a 'For Sale' sign is prominently displayed in the foreground. All elements described in the condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a two-story house with a front porch that railing, and a. There are two windows on the second floor and two windows on the first floor. The lawn appears well-maintained, and sidewalk is visible, and there 'For Sale' sign is present displayed in the foreground. The elements match in the condition are present in match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c9ad8e25ec78e2638f04020ef7f64efb382dea59
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:859b2976d37f8c46aca5b84769cb9d8f3da1a604ce45faead3d52e7e275a0eb1
+size 249162
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..0c46806f6d8bf6e92b4f6e9bab45112b6ba9d451
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a young woman taking a selfie on a sunny day. She is wearing a black baseball cap and sunglasses, reflecting the bright sunlight. Her outfit consists of a plaid shirt and a black backpack. In the background, there are other people walking by, and a woman is holding a cup of coffee. The overall style of the video is casual and candid, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "IPyA8CGk86A_32_51to232"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman, a black baseball cap, sunglasses, a plaid shirt, a black backpack, another woman holding a cup of coffee.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young woman wearing a black baseball cap, sunglasses, and a plaid shirt, with a black backpack visible on her shoulder. She is also holding a cup with a green straw, which is consistent with the description of 'another woman holding a cup of coffee' (though the cup is likely a beverage, not specifically coffee). The core elements of the description are accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young woman wearing a black baseball cap, sunglasses, and a plaid shirt. which a black backpack visible. her shoulders. The is also holding a cup of a straw liquid, which could not with the description of anotheranother woman holding a cup of coffee.' inthough it woman is not hers prop other not coffee coffee). The presence elements of the description are accurately represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..785dbfb5cdb63d75cef4964bde87f96a2bb62aa7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:859be9bbd4bdb70ba2a9c7feff6266a73c278609626daedd0b303513c870691b
+size 251789
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..1fb95d91c39266b425721ef102b740a241796bdf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car, focusing on the dashboard and steering wheel. The dashboard features a digital display with various icons and information, including the time, speed, and fuel level. The steering wheel is black and has multiple buttons on it. The car's interior is modern and sleek, with a focus on technology and convenience. The style of the video is a close-up shot, providing a detailed view of the car's interior. The video is likely meant to showcase the car's features and design."
+    ],
+    "video_ids": [
+        "-IoFtIjWSkA_23_0to198"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dashboard, digital display, icons, information (time, speed, fuel level), steering wheel, buttons? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows the dashboard with a digital display, including icons and information such as time, speed, and fuel level. The steering wheel and buttons are also clearly visible, fulfilling all the specified conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a dashboard with a digital display, which icons and information such as time, speed, and fuel level. The steering wheel and buttons are also visible visible, fulfilling the the conditions conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..66ba058342c3cb6e36dfde1612f464817676f68b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ace3dec8dafdcc14e2a224ad9e436ca426e88982d9a21aa3d9ed5a1197feef60
+size 195763
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..97385d088856655c122d62915d43ceb5cce3dd66
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment on a soccer field. The main focus is a soccer player, dressed in a vibrant red jersey with white stripes, who is in motion. He is looking up, possibly tracking the trajectory of the ball or preparing for a pass. His body language suggests he is fully engaged in the game, ready to react at a moment's notice. The background is a blur of other players, indicating the fast-paced nature of the sport. The colors are bright and the action is intense, creating a sense of excitement and anticipation. The style of the video is realistic, capturing the essence of a live soccer match."
+    ],
+    "video_ids": [
+        "qUXmW_ts-fY_9_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Soccer player in a vibrant red jersey with white stripes.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a vibrant red jersey with white stripes, which matches the description. The jersey also features the Bayern Munich logo and Adidas branding, confirming it is a soccer jersey. The player's attire and the context (on a soccer field) align with the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a soccer player wearing a vibrant red jersey with white stripes, which matches the description. The player is features a Adidas Munich logo, Adidas branding, which the is a soccer player. The player is posture and the setting ofsoc a soccer field with align with the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a0304232299898fb0b312f7f827297716afe2d7b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5289517c315c7c0502e0fe93f901a96dfc9a70921573a197edafb6eb6015c24e
+size 221197
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..2969071dd6249eeff95da0ed6e9eb46090a9a2a3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man sitting in a restaurant with a warm and inviting atmosphere. He is wearing a gray shirt and has a slight smile on his face, suggesting he is enjoying his time. The restaurant has a wooden wall with a unique geometric pattern, adding to the cozy ambiance. The man appears to be engaged in a conversation or perhaps waiting for his meal. The lighting in the restaurant is soft and warm, creating a comfortable and welcoming environment. The overall style of the video is casual and relaxed, capturing a moment of everyday life in a pleasant setting."
+    ],
+    "video_ids": [
+        "EbBmP8SsHo4_3_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man sitting at a table, wearing a gray shirt, with a slight smile.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting at a table, wearing a gray shirt, and displaying a slight smile. These elements align with the description provided, even though he is also smiling more broadly and gesturing at times, which does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting at a table, wearing a gray shirt, and he a slight smile. The elements match with the description provided. making though the is not smiling more broadly in lookinguring with one, which is not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5e423f8ce10ae02b83a15961f4870887e47ea35d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5804e7836ba45f5399307c6d4b879671eecfd62ea961ce84ee2ac43f5ed0135
+size 144704
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..0053491b1e814ce3365c5fe542d7755d7efac463
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and a black cap, wearing a black t-shirt with a white graphic on it. He is standing in front of a window with a blue frame, and there is a shelf with various objects behind him. The man appears to be speaking, and his expression is serious. The style of the video is casual and informal, with a focus on the man and his surroundings. The lighting is natural, coming from the window, and the overall atmosphere is calm and relaxed."
+    ],
+    "video_ids": [
+        "Tyy8eUzIB_s_11_0to133"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and a black cap, wearing a black t-shirt with a white graphic.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and a black cap, wearing a black t-shirt with a white graphic. These elements are clearly visible and match the description. Additional background items like shelves and figurines do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and a black cap, wearing a black t-shirt with a white graphic. The elements match consistent visible and match the description provided The elements elements like the and plantsines do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e67d4d534099dd801a8ca63c3ae50d3ddf5eb9b2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c2466a87d905bae452e9142bc44eeb277c6b3f509e81b7a35f61ce1f9c441c67
+size 109020
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..08ce860173e1ef1dd03df2e977a2e368911324ac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men engaged in a conversation on a street. The man on the right is wearing sunglasses and a plaid shirt, while the man on the left is wearing a plaid shirt and has a beard. They are standing in front of a parked car. The background features a blurred street scene with other cars and a building. The style of the video is realistic and it appears to be a candid moment captured during the day."
+    ],
+    "video_ids": [
+        "Q7NQp-xmXfY_15_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men engaging in a conversation.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men facing each other, with one speaking and the other listening, which aligns with the description of two men engaging in a conversation. The background elements (cars, trees) do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men standing each other, which one man and the other listening, which suggestss with the condition of two men engaging in a conversation. The setting and,park and trees, do not contradict the core interaction.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b1ed42d3af1429f66e1b92ca19016c4fa6b4197f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1367-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a972ce95f8759348d77096b9a1f7fe60f5788de0b726d5d45082ca839e9c54f2
+size 209819
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..9d74197a07b4343ff8db3c4f040e1a0287b59aa5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a meal being served on a green plate. The meal consists of a variety of dishes, including rice, meat, and vegetables. The plate is placed on a table, and there is a bottle of water in the background. The style of the video is a simple, straightforward food presentation, with no additional elements or actions. The focus is solely on the meal and the plate it is served on. The colors in the video are vibrant, with the green of the plate contrasting against the various colors of the food. The lighting is bright, highlighting the textures and details of the meal. The overall impression is one of a delicious, well-presented meal, ready to be enjoyed."
+    ],
+    "video_ids": [
+        "el_DEU0KwcA_1_0to110"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Green plate, rice, meat, vegetables, bottle of water? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a green plate with rice, meat (chicken), vegetables (green beans and carrots), and a bottle of water in the background. All specified objects are present and accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a green plate with rice, a,whichicken or vegetables (pe beans and red), and a bottle of water in the background. The the objects are present and match depicted in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c34ddd8eb0b846ec10ec1737eff6527cca38bb86
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e983a6c585c4a800124300fe6ae45111bfcae362ecb021cfaa837bb2195df994
+size 69268
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..9d92f4394c6a9a2df7662f428352ff061169389d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and mustache, wearing a white baseball cap and a black t-shirt. He is standing in front of a green house with white trim. The man is looking directly at the camera with a slight smile on his face. The lighting in the video is bright, suggesting it might be daytime. The man appears to be in a relaxed and casual setting, possibly in a residential area."
+    ],
+    "video_ids": [
+        "FU2Cmn1zexE_5_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and mustache, wearing a white baseball cap and a black t-shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and mustache, wearing a white baseball cap and a black t-shirt, which matches the description. The background and lighting are consistent with an outdoor setting, but these do not contradict the core description of the man's appearance.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and mustache, wearing a white baseball cap and a black t-shirt. which matches the description provided The background appears additional are consistent with the outdoor setting, and the elements not contradict the core description of the man's appearance.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..457f8786743c8e64a3a825cdc8408c279e4bbaae
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:497b4800cfee851fab5dd234103bce9813df4de99f16f76fb6390054555edce4
+size 178146
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..0ab0beb9317317060630b50a9a03068e3d3f7d4b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated depiction of a historical battle scene. The scene is set in a snowy, war-torn village with destroyed buildings and debris scattered across the ground. The main focus is on two tanks, one on the left and one on the right, moving through the village. The tanks are large and imposing, with visible treads and turrets. The snow on the ground is disturbed by the tanks' movement, creating tracks. The buildings in the background are damaged, with some showing signs of fire or explosions. The overall style of the video is realistic with a focus on historical accuracy. The lighting is natural, with shadows cast by the buildings and tanks, and the snow reflecting the light. The video does not contain any text or human figures."
+    ],
+    "video_ids": [
+        "GJ1tNkpfVA4_15_61to197"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two tanks, one on the left and one on the right.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two tanks positioned on opposite sides of a snowy, war-torn street \u2014 one on the left and one on the right \u2014 which matches the 'Object(s)' condition. Additional elements like buildings and smoke do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows two vehicles, on either sides of the snowy urban urban-torn urban. one on the left and one on the right. which align the 'Object(s)' condition described The elements like the and a do not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..58db125e4a233ad86ab2ca1d3b11dc4abe897d66
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe119908aded85e1aff60d07c6bdb09ac01e13da6cb56e3488c3e4be6e176d3a
+size 230579
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..5f5d2f43ea9b7dc910fd9a59f53aeea647f0a1e7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a stop-motion animation featuring two toy figurines. The first figure is a small, black and white striped toy resembling a classic prison uniform. The second figure is a larger, green-haired toy wearing a white tank top with a black emblem on the front and blue jeans. The larger toy is standing on a green skateboard with blue wheels. The background is a simple, colorful gradient, transitioning from blue to yellow. The style of the video is playful and whimsical, with a focus on the interaction between the two toy figurines."
+    ],
+    "video_ids": [
+        "ZXOXqYFqqGk_15_84to237"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two toy figurines - a small black and white striped prison uniform toy and a larger green-haired toy wearing a white tank top with a black emblem and blue jeans. The larger toy is also standing on a green skateboard with blue wheels.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two toy figurines as described: one in a black and white striped prison uniform and another with green hair, wearing a white tank top with a black emblem and blue jeans, standing on a green skateboard with blue wheels. The scene matches the core description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two toy figurines as described. a is a small and white striped prison uniform and the with green hair, wearing a white tank top with a black emblem and blue jeans. standing on a green skateboard with blue wheels. The scene matches the description description without any or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2b0034918722f8dcbf531112e7a76e178386a39c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f5f2cc0cca9287bd4abf4c1ff723fdd3da6557ec14db4636ff6c17558f04c11
+size 150550
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..c52a462a452c0516624229b48cd0fb3b195a818b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a cozy scene of a campfire in a forest. The fire is burning brightly, with flames licking at the logs and a pot sitting on a grate above the fire. The pot is black and appears to be cooking something. The fire is surrounded by logs, creating a warm and inviting atmosphere. In the background, a yellow sign with black text reads \"BRIDGES FREEZE BEFORE\". The sign is slightly blurred, suggesting it's in the distance. The overall style of the video is natural and rustic, capturing the essence of a camping trip in the woods."
+    ],
+    "video_ids": [
+        "9QqmWuQmk6Q_30_0to154"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Campfire, flames, logs, black pot, yellow sign? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a campfire with visible flames, surrounding logs, a black pot placed over the fire, and a yellow sign in the background that reads 'BRIDGES FREEZE BEFORE'. All specified objects are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a campfire with flames flames and logs logs, a black pot, over the fire, and a yellow sign in the background. appears 'FIDG FORCE BE BE ICE The these objects are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..414d9a77de750c91836b58813c85d7a7e0e4bc01
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:08e2056ee810bdc844525ae59d275a8675ee264f6d691e898dc014364ab58c34
+size 202091
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..6ddc631a32cab2e9540dcc80698b00bbd0fee274
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a seal swimming in the ocean. The seal is seen in three different positions, moving from the left to the right side of the frame. The seal's whiskers are clearly visible, indicating its sensitive sense of touch. The ocean water is a deep blue, and the seal's white fur contrasts with the surrounding environment. The seal's movement suggests it is hunting for fish, which it can find even in zero visibility conditions, thanks to its sensitive whiskers. The video is a close-up shot, focusing on the seal and its actions in the water."
+    ],
+    "video_ids": [
+        "mEV0vmTWwd8_17_0to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Seal? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a seal swimming underwater, which matches the 'Object(s)' condition of 'Seal'. The seal is the central focus, and its movements and features are consistent with real seal behavior and appearance. The text overlay does not contradict the object's identity.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two group in in, which ful the 'Object(s)' condition. theSeal'. The presence is the main focus of and there features and features are consistent with what-life behavior. appearance. There presenceural does not contradict the presence description condition.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..385c31e3fa3206b5b7e138a811785abb3c2b0b99
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba3d0eda81eba797c9afa784c0d082d3681db9bb6534bb207fdb8863072db534
+size 263003
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..ea3dbef55782baf0dd6256fb5a66277c75a3fa85
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a single duck swimming gracefully on a calm body of water. The duck, with its dark plumage, is positioned centrally in the frame, moving slowly from left to right across the surface. As it swims, it creates gentle ripples that spread outwards, forming concentric circles that gradually dissipate into the stillness of the water. The background features subtle reflections of surrounding trees and possibly buildings, adding depth and texture to the scene. The lighting suggests either early morning or late afternoon, casting soft shadows and enhancing the tranquil atmosphere. The overall mood is peaceful, with the duck's steady movement providing a sense of calm and continuity throughout the video."
+    ],
+    "video_ids": [
+        "818e8c2a887fa7667b8c54c171caf94280ff00446e271fa210b332e14e714be0"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A single duck with dark plumage.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a single duck with dark plumage swimming in water, which matches the description. The duck's dark feathers are clearly visible, and there are no conflicting elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a single duck with dark plumage swimming in a. which align the description of The duck is dark color and clearly visible, and there are no additional elements in contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8ab37a7c7599280929517a1d7fb772835f682443
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fa56c77d38dc95dc57e826615b4103022562ff6c5801fc71d0be51ebe6758342
+size 261284
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..196e7091cf2b0b45d90c29acff68cb094b37109d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a rusted engine, likely from a motorcycle or car, set against a grassy field with a small pond in the background. The engine is dirty and appears to be in disrepair, with visible rust and missing parts. The style of the video is realistic and documentary, focusing on the details of the engine and its surroundings. The camera angle is low, emphasizing the size and complexity of the engine. The lighting is natural, with the sun casting shadows on the engine and the grass. The overall mood of the video is somber and contemplative, as it captures the passage of time and the effects of weathering on once-functional machinery."
+    ],
+    "video_ids": [
+        "bbCVoRPv96w_26_0to122"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Rusted engine? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a motorcycle engine that appears heavily rusted, with visible corrosion on its metal surfaces. This matches the 'Rusted engine' condition described. The background and other elements do not contradict this observation.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a close-up of a motorcycle engine that appears to rusted, with visible signs and the surface surfaces. The align the descriptionRusted engine' condition as, The background, other elements, not contradict this core.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9a8187591db86bd487f5ff9d8ec03012e666c13b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1368-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eecdba784e9de4644ecd2ed98b6c7763f4622ee033f81903623204c498aba357
+size 187147
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..2e0e85e347b8f4806a48979a9de30d02297032cd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man in a blue jacket and glasses is seen interacting with a white car. He starts by pointing at the car, then opens the door, and finally, he is seen sitting inside the car. The car is parked in front of a green tree, and the scene is set during the day. The man's actions suggest that he is either inspecting the car or preparing to drive it. The overall style of the video is casual and everyday, capturing a simple moment in time."
+    ],
+    "video_ids": [
+        "5PVFHMhOD4k_46_23to244"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue jacket and glasses, a white car, and a green tree.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue jacket and glasses standing next to a white car, with green trees visible in the background. All core elements from the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a blue jacket, black, next to a white car. which a leaves in in the background. The the elements of the description are present, any.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0fe63d810f4b070d638e4541a9d8bf050878eddd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b3f5703fdf039ab7da4a7200b183fad63f5cf93c96a0407ba6c3785ce7daf884
+size 180577
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..baedde00029ac923157b83091d0e333fb8b4806c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is on a player in a black and gold uniform, who is in possession of the football. He is running towards the right side of the frame, with his teammates in pursuit. The player's uniform is clearly visible, with the number 56 prominently displayed. The teammates are dressed in similar uniforms, indicating they are part of the same team. The background is filled with the blurred figures of other players, suggesting a crowded and energetic game. The style of the video is action-packed and fast-paced, capturing the intensity and excitement of the sport."
+    ],
+    "video_ids": [
+        "96wxZ_Bw7wU_15_100to258"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a black and gold uniform with the number 56, and his teammates also in black and gold uniforms.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player wearing a black and gold uniform with the number 56, holding a football, and surrounded by teammates in matching black and gold uniforms. The description is accurately fulfilled without any conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a black and gold uniform with the number 56, and a football. and running by teammates also black black and gold uniforms. The presence is largely represented by any contradictions elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..af92cb7e78e13532c23947fa08d09868013926ad
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:20441dec51683420f114d4688c90a3d1b3b2db38fa46dfe09fb06d04bf55394f
+size 260973
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..f5ed49b6ea90c217fb4938371a6e5fb3725b817e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a young woman being interviewed on a city street. She is wearing a gray tank top and has her hair pulled back. She is also wearing white earbuds. The interviewer is holding a microphone and is standing to the left of the woman. The background of the video shows a busy city street with traffic lights, cars, and pedestrians. The style of the video is a candid interview, capturing the woman's reactions and responses to the interviewer's questions. The video is likely part of a news segment or a documentary."
+    ],
+    "video_ids": [
+        "0Zx4m2ok6D0_11_0to103"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman, an interviewer with a microphone, white earbuds.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman being interviewed outdoors, with a microphone held by an interviewer (only the hand and microphone are visible). She is wearing white earbuds, matching the description. The background includes urban elements like traffic and pedestrians, which do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a young woman being interviewed by by which a microphone held close an interviewer.part the hand and part are visible). The is wearing white earbuds, and the description. The setting includes a elements like traffic lights a, which are not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..67970d5f77b8da9fee7dd4ad700d294f61547a2b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4f4943de5987499b9c42f9ad941699de632f9e97dd6923a0203b79776f37147b
+size 184565
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..6f3c5b9a804d1e91fb46a12b14808aa493ec130c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a collection of ornate, colorful masks with intricate designs and patterns. The masks are predominantly blue and gold, with some featuring additional colors such as green and red. They are displayed in a way that suggests they are hanging from a rack or a wall, possibly in a store or a museum. The masks are arranged in a way that allows viewers to see the details of each one, highlighting the craftsmanship and artistry involved in their creation. The style of the video is straightforward and informative, focusing on the masks themselves rather than any additional context or narrative. The video is likely intended to showcase the beauty and variety of the masks, possibly for educational or promotional purposes."
+    ],
+    "video_ids": [
+        "hqnwX5ZEWvI_39_0to135"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Ornate, colorful masks with intricate designs and patterns.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features ornate, colorful masks with intricate designs and patterns, including gold embroidery, blue and green fabrics, and decorative elements like bells and pearls. These masks are displayed in a way that highlights their detailed craftsmanship, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video showcases features ornate, colorful masks with intricate designs and patterns. which vibrant and and blue and green hues, and detailed elements. swirl. rib. These masks are hanging in a way that highlights their detailed craftsmanship and which the 'Object(s)' condition described}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..72fa5f7966522d411326c71ff63217946e0a8092
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26ed20031a85702be5201073a4c7781bb1f47f8c7e185c8521c8d2a9abe96516
+size 217072
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..e6a3a8900a066fbc55c0a527a9aa92399c1dc4b6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of two fruit skewers on a table. The skewers are made of various fruits such as strawberries, pineapple, and kiwi, all dipped in chocolate. The fruits are arranged in a visually appealing manner, with the chocolate-dipped strawberries and pineapple slices being the most prominent. The skewers are placed in red bowls, which contrast nicely with the colorful fruits. The table appears to be in a kitchen, as there is a refrigerator and an oven in the background. The style of the video is simple and straightforward, focusing on the skewers and their vibrant colors. The lighting is bright, highlighting the freshness of the fruits and the smoothness of the chocolate. The video does not contain any text or additional elements, keeping the focus solely on the fruit skewers."
+    ],
+    "video_ids": [
+        "5YDR9fkW70s_31_0to174"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two fruit skewers, red bowls, strawberries, pineapple slices, kiwi, chocolate? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two fruit arrangements in red bowls, each featuring strawberries (some dipped in chocolate), pineapple slices, kiwi slices, and other fruits like cantaloupe and grapes. These elements align with the described objects, even though the arrangement is more decorative than traditional skewers. The presence of additional elements like mint leaves and chocolate-covered strawberries does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two fruit skew on red bowls, which with strawberries,red whole in chocolate), pineapple slices, andwi,, and chocolate fruits. redaloupe. possibly. The elements match with the description objects, fulfilling though the skew is not complex and a skewers. The presence of additional fruits like the leaves and the-covered strawberries does not contradict the core description but}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6496efec6cb95ad5a0262ac56282b27c6f885f7f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9c80107417a34ed55877d661031504c1c4ddd6c54d9cf33be3f4e330c3812cd
+size 148757
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..faa38d15e87be79f1e841feeb5882cc00e186592
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person with long hair and glasses is seen holding a pair of black headphones. The person appears to be in a room with a red wall and a window with a grid pattern. The headphones are held up to the person's face, suggesting that they are either about to put them on or have just taken them off. The person's expression is neutral, and they are looking directly at the camera. The overall style of the video is casual and informal, with a focus on the person and their interaction with the headphones."
+    ],
+    "video_ids": [
+        "ME0hkwMHNN0_11_0to184"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person with long hair and glasses, a pair of black headphones? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a person with long hair and glasses, and they are holding and wearing a pair of black headphones. These elements match the description in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person with long hair and glasses, and they are holding a wearing a pair of black headphones. The elements match the description provided the questionObject(s)' condition,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3def0a86ad62fdb53647cf4a48087d05a7e0cfde
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86378ca2f72250eb81a9cd056f46d4433aedf72d7d85003a77006c61be8b99ee
+size 230509
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..20f1ecdc335ba68877c7a46a8bf6aa27b65375da
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a futuristic scene with a large, glowing, blue, brain-like structure floating in the sky above a mountainous landscape. The brain-like structure appears to be made of interconnected lines and dots, giving it a digital or cybernetic appearance. The landscape below is rugged with rolling hills and a mix of green and brown vegetation. The sky is a gradient of orange and yellow, suggesting either sunrise or sunset. The overall style of the video is a blend of science fiction and nature, with a focus on the brain-like structure as the central element. The video likely explores themes of technology, intelligence, and the relationship between humans and nature."
+    ],
+    "video_ids": [
+        "I_qhIcOkDy8_0_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large, glowing, blue, brain-like structure made of interconnected lines and dots.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a large, glowing, blue, brain-like structure composed of interconnected lines and dots, which matches the description. The structure is prominently displayed and rotates slowly, allowing for clear observation of its form and details. Additional elements such as the landscape and digital artifacts do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a large, glowing, blue, brain-like structure made of interconnected lines and dots, which matches the description provided The structure is prominently displayed against is slightly, maintaining viewers a visibility of its intricate and the. The elements such as the sunset in sunset effects do not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..793d2a9194dbc06f0590406a5443f6175c4fe138
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7cb65cc75d6133e982d07a81317aefae07fcddfa55e0e9bfb13e09bce53a4868
+size 152792
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..982b3fa2a8820f6ce64ba905964517c7aff9d5c9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic city street scene. A motorcyclist, wearing a red and black jacket, is seen riding down the street, passing by a gas station and a building. The motorcyclist is in motion, weaving through traffic. The street is busy with cars and trucks, and traffic lights are visible at various points along the road. The buildings and gas station suggest an urban setting. The motorcyclist is the main focus of the video, with the city street serving as the backdrop. The video is shot in a realistic style, capturing the everyday hustle and bustle of city life."
+    ],
+    "video_ids": [
+        "Uo8C1EnX1c8_40_0to159"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A motorcyclist, a gas station, and buildings? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a motorcyclist riding on the road, a gas station on the left side of the road, and various buildings in the background, including a distinctive round building. These elements align with the specified conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a motorcyclist riding through a road, which gas station in the left side of the street, and various buildings in the background. fulfilling what tall one building with The elements align with the ' ',}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..27eb78154743baf4d66733603f35836add158dc2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1369-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89eca451b6661602ea5687d99158283584a8447708132813d708d884d76ea113
+size 248552
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..d2b0a1d9b479dcf76f9f85d1e6a1cd1869e21fe9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a close-up view of a grill where pieces of chicken are being cooked. The chicken pieces vary in size and shape, some appearing more seasoned than others. The grill grates are dark and slightly charred, indicating they have been used for cooking. Flames are visible beneath the chicken, suggesting the grill is hot and actively cooking the meat. The chicken pieces are positioned across the grill, some closer to the flames while others are further away. The camera remains stationary throughout the video, focusing on the cooking process without any noticeable panning or zooming. The lighting is natural, likely from an outdoor setting, highlighting the texture and color of the chicken as it cooks. The overall scene captures the essence of grilling, emphasizing the sizzling sound and the gradual browning of the chicken pieces."
+    ],
+    "video_ids": [
+        "f736dc81f9babfe695fe015800229b5016b8ce9b9b859d2001c691b98d701efc"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Pieces of chicken, grill grates, flames? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows pieces of chicken cooking on grill grates with visible flames underneath, matching the specified objects. The chicken appears seasoned and is sizzling on the grill, and the flames are visible through the grates, indicating active grilling. There are no conflicting elements in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows pieces of chicken being on grill grates with flames flames beneath, which the ' objects. The presence appears to, is beingizzling, the grill, which the flames are actively beneath the gapsates, fulfilling the cookingilling. The are no additional elements in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e37c90d2ff3ba2ebaa3c343bccbef35b6fa87dc1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8870f56af428905542aaac4f3dae3387eec0d0d61e49f860a53758fbb1c69d5d
+size 204588
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..1acd362f8580f9cb1413c85dcc817b986a4d5058
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a stylized, animated depiction of a nighttime scene. It features a large, full moon in the sky, surrounded by numerous stars. The moon is bright and yellow, with visible craters and shadows, suggesting it is a realistic representation of the moon. Below the moon, there is a rocky cliff with jagged edges, and a small, green plant growing on the cliff's edge. The cliff is set against a dark blue sky, which adds to the nocturnal atmosphere of the scene. The overall style of the video is cartoonish and whimsical, with a focus on the beauty of the night sky and the natural landscape."
+    ],
+    "video_ids": [
+        "L-cWapOi93Y_7_54to250"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Large, full moon with visible craters and shadows, rocky cliff with jagged edges, small, green plant growing on the cliff's edge.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a large, full moon with visible craters and shadows, as well as a rocky cliff with jagged edges. A small, green plant is also visible growing on the cliff's edge. These elements are all present and consistent with the description, even though additional characters (a turtle and a sheep) are also present in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a large, full moon with visible craters and shadows, which well as a rocky cliff with jagged edges. Additionally small, green plant is growing growing growing on the edge's edge. The elements match consistent present and match with the description provided fulfilling though the elements ora green and a frog) are present present in the scene,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..672cebc1e3d51c110369d020732cde1403071d4a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:25067f2e19be3ebbe211971fe5cc1728a6c5c542c87667606255d330de5c5f3f
+size 72994
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..4520b4d299e7ff948b14e0bf6e6ac1f9f9f9bebc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men engaged in a conversation. The first man is wearing a black jacket and has a serious expression on his face. The second man is wearing glasses and a black sweater, and he appears to be listening intently to the first man. They are standing on a porch with a white railing, and there is a tree in the background. The lighting suggests it is daytime. The style of the video is realistic and it seems to be a scene from a movie or TV show."
+    ],
+    "video_ids": [
+        "Z4zb5uepYw8_28_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men are present. The first man wears a black jacket and has a serious expression. The second man wears glasses and a black sweater, and he is listening intently.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men. The first man is wearing a dark jacket (appears black) and has a serious expression. The second man is wearing glasses and a black sweater, and he appears to be listening intently. These details align with the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men standing The first man is wearing a black jacket andlikelyars to) and has a serious expression. The second man is wearing glasses and a black sweater, and he appears to be listening intently. The details match with the description provided.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..143990307a6c49b9835ac8c4c5ec1dd53f97bc76
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0bb6e3cc01161a7c971d03b311b90c94ceb72003e20dc518fc01b16f156c9bce
+size 173473
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..76e59f4532d383c93229442e65169545326ec705
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a white rabbit is seen in a garden, interacting with a gray bucket. The rabbit starts by standing next to the bucket, then it hops onto the bucket, and finally, it jumps off the bucket and lands on the ground. The garden is lush with green plants and orange carrots. The scene is set against a wooden fence, and the overall atmosphere is peaceful and serene. The animation style is realistic with attention to detail, making the rabbit and the garden look lifelike."
+    ],
+    "video_ids": [
+        "iGAghn1Rr6A_128_0to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white rabbit and a gray bucket? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a white rabbit and a gray bucket as the central objects. The rabbit is actively interacting with the bucket, which is filled with carrots, and the scene is set in a garden. While there are other elements like carrots, tomatoes, and a bear's paw, they do not contradict the core description of the white rabbit and gray bucket.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts shows a white rabbit and a gray bucket. the main elements. The rabbit is seen interacting with the bucket, which align a with carrots, and the bucket is set in a garden-like The there are other elements like plants and plants, and a wooden, paw, they do not contradict the core description of the main rabbit and the bucket.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..42c43af4f73a09666fa273ee05fea968f161d9a6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:14a9f7046b4993714706dcb0fa36905eb7c6236df9d1b0fe05849ded9a4fc0e4
+size 179850
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..a8418c30c81cec52baa7dc19ea1935f5764bb2e5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a large white two-story house with a black shingled roof and a wrap-around porch. The house is situated on a grassy lawn with a few trees and bushes. In the first frame, the house is shown from a distance, with the porch and the lawn visible. In the second frame, the camera moves closer to the house, showing more details of the porch and the lawn. In the third frame, the camera is even closer to the house, focusing on the porch and the lawn. The style of the video is a simple, straightforward real estate video, showcasing the house and its surroundings without any additional embellishments or effects."
+    ],
+    "video_ids": [
+        "eIdttHfyzCY_20_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: House, porch, lawn, trees, bushes? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a white house with a porch, a well-maintained lawn, and surrounding trees and bushes. All the specified objects are present and accurately depicted without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a house house with a black, a lawn-maintained lawn, and several trees and bushes. The the elements elements are present and match depicted, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..22503109ddde543e87d0dbec121ab178389d7bfe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50ad76e3111587400a78569a1252e22981dbc2e4a559f556a6e3066d527eff52
+size 319277
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..76b3a0b547054402c106d8d5fb4833c8d9c4459f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are sitting on a white couch in a cozy living room. The man is holding a remote control and appears to be watching something on a television that is not visible in the image. The woman is wearing a red sweater and is holding a small elf figurine. She is looking at the man with a playful expression on her face. The room is decorated with various Christmas decorations, including a snowman and a reindeer. The overall atmosphere of the video is warm and festive."
+    ],
+    "video_ids": [
+        "HTW2QR8GUUo_1_0to147"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a woman, a remote control, a television, a red sweater, and a small elf figurine.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a man and a woman sitting on a couch, with the man holding a remote control. The woman is wearing a red sweater, and a small elf figurine is visible on the couch in front of them. Although a television is not directly visible, the presence of a remote control implies it is nearby, which is consistent with the context. All core objects mentioned are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a man and a woman sitting together a couch. with the man holding a remote control and The woman is wearing a red sweater and and she small elf figurine is visible in the table. front of them. The there television is not directly visible, the presence of the remote control suggests that is in, fulfilling is a with the ' of The the elements are in present in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0048a6247fb78e3db5d29ffaef7d4299556ca75a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba60a3d5afad7aade5d6e6b8df3e22cf417d2772d4b1afece8eb96f6f8fe8297
+size 145895
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..5760b17deab00b2d7181006c05c07b1158067b3b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video opens with a close-up shot of a small, light-colored animal, likely a pika or similar rodent, moving swiftly across a rocky terrain. The animal's fur blends well with the gray and white rocks, making it initially difficult to spot. As the camera pans out, we see the animal running along a path lined with large boulders and patches of green grass. The background reveals a rugged mountainous landscape under a clear blue sky, suggesting a high-altitude environment. The camera follows the animal as it navigates through the uneven ground, occasionally pausing to sniff the air or look around. The scene captures the natural agility and alertness of the creature in its habitat, emphasizing the stark beauty of the alpine setting."
+    ],
+    "video_ids": [
+        "b10a8db608c66e1196bea95aaa755740d60d381173bf3297586b022c03150d91"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small, light-colored animal (likely a pika or similar rodent).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video consistently shows a small, light-colored animal moving through a rocky, grassy environment, which matches the description of a pika or similar rodent. The animal's size, coloration, and behavior are consistent with the expected characteristics of such an animal.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts shows a small, light-colored animal that across a snowy, snowyy terrain with which align the description of a pika or similar rodent. The animal's size, color,, and the align consistent with the characteristics characteristics of a an animal in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0972d9e672be3bd3856dc4e2bfd745b4646d530d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4336159e467a3680844372c3b93071b7758265ba8ba46a7b45737e6b857d4758
+size 281057
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..2a1394d634df68ca3d64fff5d468c4d5579e0546
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a group of people walking along a wooden boardwalk that hugs the edge of a cliff. The boardwalk is situated above a rocky shoreline, where the waves crash against the rocks, creating a dynamic and powerful scene. The sky is overcast, casting a soft light over the scene. The people are dressed in casual clothing, some carrying backpacks, suggesting they are out for a leisurely walk or hike. The boardwalk is made of wooden planks, and it appears to be sturdy and well-maintained. The overall style of the video is naturalistic, capturing the beauty of the landscape and the casual, unhurried pace of the people walking along the boardwalk."
+    ],
+    "video_ids": [
+        "SokSuP4Y9Zw_31_84to282"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A group of people, wooden boardwalk, rocky shoreline, waves, sky? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a group of people walking along a wooden boardwalk, a rocky shoreline with waves crashing against it, and an overcast sky. All specified elements are present and accurately depicted without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a group of people walking on a wooden boardwalk. with rocky shoreline, waves crashing against it, and a overcast sky. The these elements are present and contribute depicted, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f85482440c575efe92de767db62932169ca11c47
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-137-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:58d996005c9a801339f25c8ff39bb4f51a2c9c44c7baefde1676815d77ffba1c
+size 255325
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..561372daafeaebbc022a2a81ce3a218cf0e96c8d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a silver sports car with the number 69 on its side, driving on a race track. The car is sleek and shiny, with a black roof and black rims. The track is wide and paved, with a red barrier on the side. In the background, there are other cars parked, and a grassy field can be seen. The sky is clear and blue, suggesting a sunny day. The car is moving at a high speed, as indicated by the blurred background. The style of the video is dynamic and action-packed, capturing the thrill of racing."
+    ],
+    "video_ids": [
+        "FGIdYeEKppM_3_0to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A silver sports car with the number 69 on its side, and other parked cars.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a silver sports car prominently featuring the number 69 on its side, consistent with the description. Additionally, other parked cars, including what appear to be race cars or similar vehicles, are visible in the background. These elements align with the 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a silver sports car with in the number 69 on its side, which with the description. Additionally, there parked cars are which a appears to be race cars, sports vehicles, are visible in the background, The elements align with the descriptionObject(s)' condition described contradict.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..645bf02450ad776950931703ae741bf539c52630
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dda9cbacedafc644a4074f11b307adc1e4795b06e9254cfe7a020947d032bdc4
+size 198963
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..135b3c73c9bf08a01e71a5218773f1c0cc5e5e8a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man standing next to a red Tesla car on a rural road. The man is dressed in casual attire and appears to be interacting with the car, possibly opening the door or adjusting something on the vehicle. The car is parked on the side of the road, and the surrounding environment is lush and green, suggesting a countryside setting. The style of the video is realistic and appears to be a candid moment captured during a road trip or a test drive. The focus is on the man and the car, with the background providing context to the location."
+    ],
+    "video_ids": [
+        "WkTaWqsSzBk_40_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a red Tesla car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man standing next to a red Tesla car, which matches the core description. The surrounding environment and additional elements (like the landscape and the 'lovecars' watermark) do not contradict the presence of the man and the car.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man interacting next to a red Tesla car. which align the ' description of The man environment, additional elements,like the second and another secondHosear' text) do not contradict the main of the man and the red.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c35bc39ba7994b3a280cf6c4e8accbbf9e85b85a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd6e24254794c12ddcd041efd40b8bd147edfa0b4d8083aa708e6abad8e87333
+size 197434
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..dba5ac7f3a6f278fa96a7a9c4013436b9c7e8657
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features three men with tattoos, standing in front of a colorful mural. The men are shirtless, showcasing their tattoos, and appear to be in a relaxed, casual pose. The mural behind them is vibrant and features various designs and patterns. The style of the video is artistic and expressive, capturing the unique personalities and styles of the men. The focus is on the tattoos and the mural, creating a visually striking image. The video likely conveys a sense of individuality and self-expression."
+    ],
+    "video_ids": [
+        "IVnnWDXNWVE_13_0to104"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three men with tattoos, standing shirtless.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three shirtless men with extensive tattoos covering their bodies and faces, standing in a row. This matches the core description of 'Three men with tattoos, standing shirtless.' Additional elements, such as the background decorations and jewelry, do not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three menless men with visible tattoos on their upper. chests. standing in front line. The matches the description description of theThree men with tattoos, standing shirtless.' The elements such such as the colorful and and the, do not contradict the description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0358a4c2ac63457972dfc9fbf4fc37203c8a7e46
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb49777c33183a957c1b3d38c6aa176c5e7a1b4b506f134c7ebb50818d2a0968
+size 186174
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..8d0981aa65bca5119891dc1f0aee25046345cf2f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man named James Gregory, who is identified as a manager. He is seen in a bar setting, with a blurred background that suggests a lively atmosphere. The man is dressed in a white shirt and appears to be in deep thought or possibly listening intently to someone off-camera. The lighting in the bar is warm and inviting, with soft lighting that highlights the man's features. The overall style of the video is candid and informal, capturing a moment in the man's life in a relaxed and unposed manner."
+    ],
+    "video_ids": [
+        "CVTsuArB9ZQ_14_123to338"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man named James Gregory, dressed in a white shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man who appears to be James Gregory, as indicated by the on-screen text. He is wearing a white shirt with a checkered pattern, which matches the description. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing is to be dressed Gregory, as he by the description-screen text. He is dressed a white shirt, a collarered pattern, which align the description. The setting is lighting elements in not contradict the description description.\"\n}<|im_end|>\n addCriterion",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..779c0902a1312f8b8101ca447d9e166995266aca
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e372811ed8b3b55df625fe8e8058270f112c1520e4ecc31359b10c146990902d
+size 153157
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..4d6bc1d22e334e1e22af1af5a03a7e2cdc96f381
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a soccer player in action, wearing a blue jersey with white stripes and a red and white logo on the front. He is seen in three different positions, running across the field with intensity and focus. The background is a blur of spectators, indicating a live game scenario. The player's movements and the crowd's presence create a dynamic and energetic atmosphere."
+    ],
+    "video_ids": [
+        "siwD6kMjnIk_28_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A soccer player wearing a blue jersey with white stripes and a red and white logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a blue jersey with white stripes on the shoulders and a red and white logo on the chest, which matches the description. The jersey also features the sponsor 'VEOLIA' and other minor details like armbands and tattoos, which do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two soccer player wearing a blue jersey with white stripes and the sleeves and a red and white logo on the chest. which matches the description provided The player also has a word 'JGAIA' on the details details that thembands and socks, which are not contradict the core description.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..379129357781780e64cbf246166eee6c9336565a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8eed798ad5b20984cdb32b98059cd884e4e05338975e548f898a350877aa7a5c
+size 215295
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..1d63482c6bfa571aa010750eca40617e4e60be14
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a Canada goose swimming gracefully on a calm body of water. The goose, characterized by its distinctive black head, white face, and brownish-gray body, is seen moving from left to right across the frame. As it swims, the goose dips its head into the water, creating ripples that spread outward. The water's surface is dotted with small green leaves and twigs, adding a touch of natural beauty to the scene. The background features a mix of blue sky and water, with some green foliage visible on the left side of the frame. The overall atmosphere is peaceful and tranquil, highlighting the natural behavior of the goose in its aquatic environment."
+    ],
+    "video_ids": [
+        "a1cee2232b98b7b052ac8ebf766881e0716c5eacaebc0b43e736a47b148f03c8"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A Canada goose? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bird with the distinct black head and neck, white cheeks, and brown-gray body feathers characteristic of a Canada goose. It is swimming in water, which aligns with the typical behavior of this species. The visual details match the description of a Canada goose.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a Canada that the characteristic features head, brown, brown face, and brown body body that characteristic of a Canada goose. The is swimming in a, which iss with the typical habitat of a species. The presence elements, the description of a Canada goose,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0390233256ff36dff51993e1379251bc2df023dd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cfbd038492a38a87331316ccfd372ead9269903db3b95291906cc127147a3d75
+size 390216
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..e58d70e05366ade40988ece8fac900bc5ebb05cf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a black tire with a red logo on the center cap, which is a part of a blue truck. The tire is positioned on a grassy field, and the truck is parked nearby. The style of the video is a straightforward, real-life depiction with no additional effects or filters. The focus is on the tire and the truck, and there are no other significant objects or actions in the video. The lighting and composition suggest that the video was taken during the day under natural light."
+    ],
+    "video_ids": [
+        "x0fWLc9w_Mk_19_0to104"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black tire with a red logo on the center cap, a blue truck? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a black tire with a red logo on the center cap, mounted on a blue truck. The tire's tread and sidewall text are visible, and the truck's blue body is partially shown in the background. The scene is set on grass, which does not contradict the description. All core elements described are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close-up of a black tire with a red logo on the center cap, which on a blue vehicle. The tire and design pattern theall are are visible, and the red's blue color is partially seen in the background. The presence is set outdoors a, which is not contradict the description. The elements elements of in present in match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..41f221a686b741f4677eacba2c99c0be437a1e22
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:08c4243d19cff851b511d3d8e0a53c91a7e0a1efea45d3ecb59f044624dab29a
+size 101655
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..dfa2edd2d4beb72381bb24fcaf6ff9984f35510c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a cooking tutorial featuring a chef in a professional kitchen. The chef, wearing a white chef's coat and glasses, is preparing a large roast with a green herb crust. He is using a knife to trim the roast and is holding it up to show the herb crust. The kitchen is well-equipped with various cooking utensils and ingredients, including a cutting board, a knife, and a bowl of fresh herbs. The chef is standing in front of a stove and a sink, indicating that he is in the midst of cooking. The style of the video is informative and instructional, with the chef providing step-by-step instructions on how to prepare the roast. The video is likely aimed at home cooks looking to learn new cooking techniques."
+    ],
+    "video_ids": [
+        "LrjLz0F292E_5_17to193"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chef, large roast with a green herb crust, knife, cutting board, bowl of fresh herbs? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chef in a kitchen setting holding a large roast with a green herb crust, standing behind a cutting board. A knife is visible on the counter, and there is a plate with fresh herbs nearby. All the specified objects are present and clearly visible, matching the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chef in a kitchen environment, a knife roast with a green herb crust. using on a cutting board. The knife is being in the cutting, and there is a bowl of fresh herbs in. The elements elements elements are present and match visible in fulfilling the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0765630ea5fddf3c226afa19c72195ef7a62a96d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1370-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:39aad0e6fbc9fc7447eec1913b6a7a7044f0fef7fb63201a1e1855decbc7d33f
+size 244482
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..80c2a02e53b357af55e1aae158d318103ccb0af4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in front of a large screen displaying a silhouette of a leopard. The man is wearing a black shirt and glasses, and he appears to be speaking or presenting. The style of the video is a combination of a live event and a visual aid, with the man as the central figure and the leopard silhouette as the backdrop. The focus is on the man and his interaction with the audience, while the leopard silhouette adds a visual element to the presentation. The overall atmosphere of the video is professional and informative."
+    ],
+    "video_ids": [
+        "3wBGeYU1IHM_80_0to101"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man wearing a black shirt and glasses, who is speaking or presenting.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black shirt and glasses, standing in front of a backdrop with a panther silhouette. His mouth is moving, suggesting he is speaking or presenting, which aligns with the 'Object(s)' condition. The background does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black shirt and glasses, standing in front of a screen that an largether illustration. The posture is open, suggesting he is speaking or presenting. which aligns with the 'Object(s)' condition described The presence and not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..504067af6f9e5c2c118125b01cf09eea8b15022f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26df759657d2cd21c23a0245bc34268dadb627a1b3a1e40a93c59e75e12b7a9e
+size 141071
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..2f07b9e755d3d2f229bed925b44c2479f3ba8006
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a kitchen, preparing food. She is standing at a counter, which is filled with various kitchen items such as bowls, spoons, and a blender. The kitchen is well-lit, with natural light coming in through the windows. The woman is wearing a blue shirt and is focused on her task. The overall style of the video is casual and homey, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "BbUaI4ZejVk_24_39to215"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a counter, bowls, spoons, a blender? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman in a kitchen setting, standing at a counter. She is using bowls and spoons to transfer a substance, likely frosting or batter, into a piping bag. A stand mixer (which can be considered a type of blender or mixer) is visible on the counter. All core objects mentioned in the condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a woman in a kitchen setting, standing at a counter. There is interacting a, aons, prepare ingredients yellow from likely ingredients, a, into a blender bag. The blender mixer iswhich can be considered a type of blender) food) is also on the counter, The the elements mentioned in the condition are present and correctly depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c74ef3bd220fb86f840ce9abcb78f872b0c312d8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a5423d7d0e81c9d4ad95ecb2dcecdeed158ca5f1c97941c0b5285dad5c192da
+size 142988
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..74d93ace533e13aeebf8142580a4fa994cd3c79c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in a kitchen with red cabinets. He is wearing a black shirt and appears to be speaking. The kitchen has a modern design with a countertop and a clock on the wall. The man is the main subject of the video, and his actions suggest that he might be presenting or explaining something. The style of the video is straightforward and seems to be focused on the man and his surroundings."
+    ],
+    "video_ids": [
+        "0z52UtGiOYA_8_104to259"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man wearing a black shirt standing in the kitchen and appearing to speak.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black shirt standing in what appears to be a kitchen or workshop setting with red cabinets. He is actively speaking and gesturing, which aligns with the description. The background elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black shirt standing in a appears to be a kitchen. a setting. red cabinets and He is gest gest and gesturing with which aligns with the description of The presence includes, not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3e926ee4cd17a42739a2f43d23882589be1fd668
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5d04293d8af9cab74d46db634cb4ce211b68960ec8c80146c8a69b969847fb9
+size 118023
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..210ab1b71c31a7db2f3c57a7fe02d89882b36554
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a stylish outfit, including a blue shirt and sunglasses, standing next to a silver sports car with the door open. The car is parked on a road with a clear sky in the background. The man appears to be getting into the car, suggesting a sense of motion and activity. The overall style of the video is sleek and modern, with a focus on the man and the car, creating a sense of luxury and sophistication."
+    ],
+    "video_ids": [
+        "kYimo6JZZkc_8_0to103"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a stylish outfit (blue shirt, sunglasses), a silver sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue shirt and sunglasses, leaning against a silver sports car with the door open. The man's outfit and the car match the description, and no elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue shirt and sunglasses, which against a silver sports car. a door open. The man's outfit and the car match the description provided fulfilling there additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1522a2d2b3339f9a4b272a2fb2dc548aeb080e62
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:498857312a7bb6adb36cfeced1834c90b2b7fc675a35596276f7cdce808ac3f7
+size 190003
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..20b6881b3d003b797930032e42e986e6e5a2ef12
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a blue Volkswagen car parked on a street. The car is the main focus of the video, and it is captured from the rear view. The car is stationary, and there are no people or other vehicles visible in the video. The car's license plate is clearly visible, and it reads \"BL-44H\". The car is parked in front of a building, and there is a street lamp visible in the background. The video is a simple, straightforward shot of the car, with no additional action or movement. The style of the video is straightforward and unembellished, focusing solely on the car and its immediate surroundings."
+    ],
+    "video_ids": [
+        "JFGXvNYH5rM_0_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Blue Volkswagen car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a blue Volkswagen car, specifically a rear view of a modified sedan with custom wheels and a lowered suspension. The car's color, make, and model are consistent with the description, and no elements contradict this core identification.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a blue Volkswagen car, which a model view of a Volkswagen Volkswagen. aizations and a license suspension. The car's license, model, and model are consistent with the description of and there conflicting contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a39d0c038daa45a196b362bcd45d92e4b2ce4fbf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4605f2272b0646c1fb4e80c49b00d989226ba0ae9b2b797c6cf043dd7ef9be2
+size 155319
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..51001a64964042bb332bd1f5f6f946a42e00887d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a playful and whimsical scene featuring a large orange pumpkin with carved eyes and a small pink pig figurine. The pig is standing next to the pumpkin, appearing to be peeking out from behind it. The pumpkin is placed on a green leaf, and there is a small pile of pumpkin seeds nearby. The background is a simple, neutral surface that does not distract from the main subjects. The style of the video is light-hearted and fun, with a focus on the interaction between the pig and the pumpkin. The pig's position next to the pumpkin and its peeking action suggest a sense of curiosity and playfulness. The pumpkin's carved eyes give it a friendly and welcoming appearance. The pumpkin seeds add a touch of realism to the scene, reminding the viewer of the natural process of growing pumpkins. Overall, the video captures a charming and delightful moment of interaction between a pig figurine and a carved pumpkin."
+    ],
+    "video_ids": [
+        "sf9A3MbaAwc_25_21to235"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large orange pumpkin with carved eyes and a small pink pig figurine, a small pile of pumpkin seeds? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a large orange pumpkin with carved eyes and a small pink pig figurine (Peppa Pig), along with a small pile of pumpkin seeds that appear to have been scooped out from the pumpkin. The core objects described are present and accurately depicted, even though there are additional elements (like tomatoes and another pumpkin) in the background, which do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a large orange pumpkin with carved eyes and a small pink pig figurine.whichppa Pig) along with a small pile of pumpkin seeds. are to be been carvedoped out from the pumpkin. The elements elements in in present and match depicted in fulfilling though the are additional elements likelike the and leaves pumpkin) that the background, which do not contradict the main.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..749e48ec6c8736a17538280703e0e5a6cd113adc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6463cd5a03ff959a6c6883a78750629a3f4d3529ccc81bad1354f4cd77525099
+size 162523
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..aa5974418062de333f69cb423f72786ced090e3e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men posing for a photo in a room. The man on the left is wearing a white hoodie with Chinese characters on it and glasses. The man on the right is wearing a black shirt and has a bald head. They are both smiling and appear to be in a good mood. The room they are in has a beige wall and a picture hanging on it. There is also a desk with a computer monitor on it in the background. The style of the video is casual and friendly, capturing a moment of camaraderie between the two men."
+    ],
+    "video_ids": [
+        "71-UD_l7nfM_13_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men posing for a photo; one wearing a white hoodie with Chinese characters and glasses, the other wearing a black shirt and having a bald head.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men posing together. The man on the left is wearing a white hoodie with visible Chinese characters and glasses, and the man on the right is bald and wearing a black shirt. These details match the description provided in the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men posing for, One man on the left is wearing a white hoodie with Chinese Chinese characters and glasses, while the man on the right is wearing and wearing a black shirt. The details match the description provided, the questionObject(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c76752c51c86eb2865912e274fc49d626e2a5dfa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:82fa1a800bcdf13d6da38b1b6d790913ee9d77aeb0cf73f962e6fed3f974918e
+size 136331
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..28c1080d0a802157ad0cf7a967e2f438926b9281
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a close-up scene set on a tiled countertop. A clear plastic container filled with water is placed centrally on the surface. On top of the container, a piece of paper with a black and white illustration of a cat wearing a suit and tie is positioned. The cat's image is detailed, showing it standing on a striped surface, possibly a bar or a similar patterned area.  A hand enters the frame from the right side, holding the paper steady above the water. The hand then gently lowers the paper into the water, causing ripples to spread across the surface. As the paper is submerged, the cat's image becomes partially visible through the water, creating an interesting visual effect where the illustration appears to float or be partially submerged. The hand continues to hold the paper until it is fully immersed, allowing the viewer to see the interaction between the paper and the water more clearly. The background remains consistent throughout, featuring a green cutting mat and some other items on the counter, but"
+    ],
+    "video_ids": [
+        "7abaa3a0590c575070952851370fa3507518f4d5c90fe108d9cfa1c6cb0babd3"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Clear plastic container with water, a piece of paper with a black and white cat illustration, a hand holding the paper.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a clear plastic container filled with water, a piece of paper with a black and white illustration of a cat (or dog-like animal), and a hand holding the paper above the water. All core elements described are present and consistent with the video content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a clear plastic container filled with water, a piece of paper with a black and white cat of a cat,d a), figure), and a hand holding the paper. the water. The these elements of in present, match with the description content.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8bb1a672267e39ed821ac1ff79c11b0fa66a6660
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1371-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f9fa4922208f6c3fe458cffa42a191d7c03296398cfeeb5a2deed069f5f20b7
+size 181753
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..a3117609d0672fd745f45cb382103b30856545cc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young man with a beard and mustache, wearing a red baseball cap and a black and white checkered shirt. He is standing against a blue background with vertical stripes. The man appears to be speaking or gesturing with his hands, suggesting that he might be giving a presentation or explaining something. The style of the video is casual and informal, with a focus on the man's facial expressions and body language. The lighting is bright and even, highlighting the man's features and the details of his clothing. The background is simple and uncluttered, allowing the viewer to focus on the man and his message."
+    ],
+    "video_ids": [
+        "4YDrUlmIuDE_3_0to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man with a beard and mustache, wearing a red baseball cap and a black and white checkered shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man with a beard and mustache, wearing a red baseball cap and a black and white checkered shirt, which matches the description. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man with a beard and mustache, wearing a red baseball cap and a black and white checkered shirt. which matches the description provided The background and additional elements are not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8d795494fc7b8442f56e0e54e913a1ce95ae9b95
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c2b73406cfca6d26bcb23a508dc8d7e4392e5ddbf6d10902c055c862f1b3ed2d
+size 196597
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..d764c59677aaacda4b542d9a94fdb6214079da36
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in front of a blue display case filled with various cleaning supplies. He is holding a large green bottle of cleaning solution, which is prominently displayed in the center of the image. The man is dressed in a white shirt with orange stripes, and he appears to be presenting or demonstrating the product. The background is a simple blue wall with white text that reads \"CleanItSupply.com\", indicating that the video is likely an advertisement or promotional material for the website. The overall style of the video is straightforward and informative, with a focus on the product being showcased."
+    ],
+    "video_ids": [
+        "CPkyE3gj2r0_0_0to183"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a blue display case, various cleaning supplies, a large green bottle of cleaning solution, a white shirt with orange stripes.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man in a white shirt with orange stripes, standing in front of a blue display case filled with various cleaning supplies. He is holding a large green bottle of cleaning solution. All elements mentioned in the description are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man standing a white shirt with orange stripes, standing in front of a blue display case. with various cleaning supplies. The is holding a large green bottle of cleaning solution, The elements in in the description are present and accurately depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c602c52b8aeab3156c20e06915e2ef2f1f4298d3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b42d3a0797c2f86c326c325d9d1d8608d0dfb1a3cb584e0f0e97221c358d3624
+size 247672
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..f75e1a53040f330a1e15e64ac90db8940419f2b8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a collection of potted plants arranged on a wooden surface. The plants vary in size and type, with some having spiky leaves and others with more rounded foliage. The pots themselves are also diverse, with different colors and textures, including terracotta, black, and white. The plants are positioned in a way that creates a visually appealing arrangement, with some pots placed in the foreground and others in the background. The wooden surface provides a natural and rustic backdrop for the plants, enhancing their colors and textures. The overall style of the video is simple and straightforward, focusing on the beauty of the plants and the pots in which they are housed."
+    ],
+    "video_ids": [
+        "AXRSuyf49Vk_22_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Potted plants and various pots? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features multiple potted plants, including succulents and cacti, arranged in various pots of different sizes and materials (terracotta, black plastic, etc.). These elements align with the 'Object(s): Potted plants and various pots' condition, with no conflicting elements present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features p potted plants, including aulents and aacti, which on various pots. different colors and colors.cerracotta and black).). and.). The elements clearly with the 'Object(s)' Potted plants and various pots' condition, as the conflicting additional present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8943342aa71bc76987a54865ac48108168b96746
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5b057303a2d34acbbce3f2640837e153717ddc582adafbf7b6a83a2d621cf6f
+size 84857
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..10c34f62b05baf57caa2b379c02fc17d72374ace
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two medical professionals, likely doctors, examining a patient's MRI scan displayed on a monitor. They are wearing surgical caps and are focused on the image, which shows a detailed view of the patient's head and brain. The doctors are using a pen to point out specific areas of interest on the scan, indicating a collaborative discussion about the patient's condition. The setting appears to be a medical facility, and the overall style of the video is informative and professional, likely used for educational or diagnostic purposes."
+    ],
+    "video_ids": [
+        "FzoMvujgQd0_10_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two doctors, a monitor displaying an MRI scan.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two doctors wearing surgical caps and scrubs, standing in front of a monitor displaying an MRI scan of a brain. The scene matches the core description without contradiction, and additional elements (like the background wall or electrical outlets) do not interfere with the primary objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two individuals in surgical caps and masksubs, and in front of a monitor that an MRI scan of a brain. The presence align the description description of any or as the elements suchsuch the doctors and and the equipment) do not conflict with the main focus being}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c7068ebfbe2362195649a647ad9485953c22e057
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06f3193e4b5cbb3709c3c98e690ce00ddbc0cdd00fd8d008d9c4939630157f60
+size 130916
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..e54a7b788b3aab686e94a6de3c55dcd836fa3e56
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a car engine on display at a convention or show. The engine is disassembled, revealing its intricate components and parts. The style of the video is informative and educational, showcasing the inner workings of the engine. The engine is mounted on a stand, and the camera captures the details of the engine from various angles. The background is blurred, focusing the viewer's attention on the engine. The video is likely intended for car enthusiasts or those interested in automotive engineering."
+    ],
+    "video_ids": [
+        "6DmbTlXUpJQ_31_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A car engine (disassembled)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a car engine that appears to be disassembled, with various internal components visible and mounted on a display stand. The engine is the central focus, and while there are people and background elements, they do not contradict the description of a disassembled car engine.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close engine that appears to be disassembled, with various components components visible. some on a stand stand. The presence is detailed central focus, and the there are some and other elements, they do not detr the core of the carassembled car engine.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..897ebe18e90b6daedc2c88f820289ed6d83bebd6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3cb91de0f0de651889a0606f692ea3016d97f99c345d3179a6716c7ba36ee2e
+size 165702
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..91f646c12dc6f7df20f41cfbb511a2dd609b259d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen preparing a pie in a kitchen setting. The person is pouring a yellow liquid, likely a filling, from a bowl into a pie crust. The pie crust is placed in a pie dish, which is on a countertop. The person is using a spoon to transfer the liquid. In the background, there are bottles and a bowl, suggesting that the kitchen is well-equipped for cooking. The style of the video is a simple, straightforward cooking tutorial, focusing on the process of making a pie."
+    ],
+    "video_ids": [
+        "UHqHSVgOXI8_11_22to191"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person, pie crust, pie dish, bowl, spoon, bottles? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person pouring a yellow filling into a pie crust inside a pie dish, using a bowl and spoon. Bottles are visible in the background, fulfilling all the listed object conditions without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a person pouring a yellow liquid into a pie crust that a pie dish. which a bowl. a. Theles are also in the background, which the the conditions objects conditions.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e706760659da581c1612d89f84cdde1272d94018
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3e76dd3a63a2be8560b71b7dafe9b0fa2e5cabefede433978edbd96a9a7fa13
+size 115527
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..ddd0e0acaefabde68a6fd48675ade765abde19dc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is on a player in a black and gold uniform, who is in possession of the football. He is running towards the right side of the frame, with his teammates in pursuit. The player's uniform is clearly visible, with the number 56 prominently displayed. The teammates are dressed in similar uniforms, indicating they are part of the same team. The background is filled with the blurred figures of other players, suggesting a crowded and energetic game. The style of the video is action-packed and fast-paced, capturing the intensity and excitement of the sport."
+    ],
+    "video_ids": [
+        "96wxZ_Bw7wU_15_100to258"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a black and gold uniform with the number 56, and his teammates also in black and gold uniforms.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player wearing a black and gold uniform with the number 56, holding a football, and surrounded by teammates in matching black and gold uniforms. The description is accurately fulfilled without any conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a black and gold uniform with the number 56, which a football. and running by teammates in similar black and gold uniforms. The setting is largely represented by any contradictions elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8ff039013681bbb9e80edf34207e74520032b69c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a4902b99940cbc5a474af5776ecbb7f454077f1190d7793df2c704b3e90fdbeb
+size 263410
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..024728752b438ca8edea03edf0828c3a089c4355
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man with a beard and a plaid shirt is sitting on a couch in a living room. He is gesturing with his hands as he speaks, indicating that he is engaged in a conversation or explaining something. The living room has a warm and cozy atmosphere, with a lamp providing soft lighting and a window in the background. The man appears to be the main subject of the video, and his actions and expressions suggest that he is the one speaking or explaining. The overall style of the video is casual and informal, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "6RxzbgyDI1s_20_0to113"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and a plaid shirt sitting on a couch.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and a plaid shirt sitting on a couch, which matches the core description. Additional elements like background furniture and lighting do not contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and a plaid shirt, on a couch. which align the description description. The elements like the details and lighting do not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..32ad65e9cbe1e96930797aa3ec314cbd6876fe4b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1372-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:abc3e2601d5b51fbb2eaca4e50fe90197a52ea2067136395d385532ce72e5036
+size 323242
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..2bd035d30b0c9d173d3b92572cccd32d5c08e8aa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a dessert being eaten. The dessert is a slice of cheesecake, which is being scooped up with a spoon. The cheesecake is placed on a plate, and there are pieces of fruit, including strawberries and kiwi, arranged around it. The plate is on a table, and there is a bowl in the background. The style of the video is a simple, straightforward food video, focusing on the dessert and the process of eating it. The lighting is bright, and the colors are vibrant, highlighting the textures and colors of the food. The video does not contain any text or additional elements, and the focus is solely on the dessert and the act of eating it."
+    ],
+    "video_ids": [
+        "Ml5AAM_O6lE_46_0to163"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Cheesecake slice, spoon, strawberries, kiwi, plate, bowl? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a cheesecake slice, a spoon scooping from it, strawberries, kiwi, a plate, and a bowl in the background. All specified objects are present and identifiable in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as showing showing a cheesecake slice, a spoon,oping the it, strawberries, kiwi, and plate holding and a bowl in the background. The these objects are present and identifiable in the scene.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1b16df0af377673767d17054bd080650cbf0788c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f063a8bb1456522b87a866fd0df412cbfa4dd944adafbadcdea7faa033e5722c
+size 180398
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..e9e57dc9cbe1570850ed73d5613841994b60be9e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a car review showcasing the interior of a Mercedes-Benz vehicle. The style of the video is a close-up, interior shot with a focus on the car's dashboard and steering wheel. The car's interior is luxurious, featuring a sleek design with a large touch screen display, a digital instrument cluster, and a variety of buttons and controls. The steering wheel is equipped with a Mercedes-Benz logo and has multiple buttons for controlling various functions. The car's interior is well-lit, with the sunlight streaming in through the windows, highlighting the car's features and design. The video is likely intended for car enthusiasts and potential buyers, providing a detailed look at the car's interior and features."
+    ],
+    "video_ids": [
+        "JlQH9Jm_eh8_21_0to144"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dashboard, large touch screen display, digital instrument cluster, buttons and controls, steering wheel with Mercedes-Benz logo and control buttons.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the dashboard, a large touch screen display, a digital instrument cluster, various buttons and controls, and a steering wheel prominently featuring the Mercedes-Benz logo along with control buttons. All specified elements are accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a interior of a large touch screen display, digital digital instrument cluster, buttons buttons and controls, and a steering wheel with featuring the Mercedes-Benz logo and with control buttons. The these elements are present depicted in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..24badfd8d16077eee33750307848e4d4f696e488
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9a0a6236cc085383a386c658891939ca22da76b25db6884b383ae64a0da45d0
+size 268172
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..4fe22953b3b8b859c97bc0fc70d39a88798fb845
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a scene featuring two dogs interacting with a large mirror placed outdoors on a concrete surface. The setting is a natural environment with lush greenery and trees in the background, suggesting a rural or semi-rural area. The mirror has a wooden frame and reflects the surroundings, including the dogs.  Initially, one dog, with a light brown coat and white markings, approaches the mirror and sniffs at it curiously. Its tail is raised, indicating interest or curiosity. As the dog continues to investigate, another dog, also light brown but with a slightly different coat pattern, joins from the right side of the frame. This second dog appears more cautious and hesitant, sniffing the ground near the mirror before looking at its reflection.  The first dog continues to explore the mirror, moving closer and sniffing around it. The second dog remains more distant, observing the first dog's behavior. Both dogs exhibit typical canine curiosity, with their tails wagging slightly as they interact with their reflections. The camera"
+    ],
+    "video_ids": [
+        "42018f3b7c565bb9b09718a384becebe83729c0ed278c17ec988b87ea4d75ee3"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two dogs, a large mirror, lush greenery, and trees.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing two dogs interacting with a large mirror. The background features lush greenery and trees, as described. Although a chicken briefly appears, it does not contradict the core description and is an acceptable additional element.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as depicting showing two dogs, with a large mirror. The lush includes lush greenery and trees, which described. The the cat is appears in it does not contradict the core description and is not additional addition element.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a992e3528129f47aa33ebc2c85fa1488f1f88160
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d32fa115a7cc7f83dbac855454dd6e3d6616551184123a6136189eee8c5186c7
+size 192994
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..44ef4bc38f2f7365ae764b7e373389459e807b8b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen enjoying a meal at a restaurant. She is holding a large bowl filled with a delicious-looking dish, possibly a noodle dish or a salad, topped with a generous amount of nuts. The bowl is white, contrasting with the colorful food inside. The woman is seated at a table, and the restaurant has a lively atmosphere with other patrons visible in the background. The lighting is warm and inviting, creating a cozy ambiance. The video captures the woman's enjoyment of her meal, the vibrant colors of the food, and the bustling restaurant environment."
+    ],
+    "video_ids": [
+        "H030kvQqrDA_38_0to162"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a large white bowl filled with colorful food, nuts on top of the food, tables and chairs in the restaurant.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman holding a large white bowl filled with food that appears to have nuts on top. The setting is a restaurant with tables and chairs visible in the background. All core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman eating chop fork white bowl filled with colorful that appears to be a on top. The setting is a restaurant with tables and chairs visible in the background. The elements elements of in present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3d0d0d05d61ce882d0ba5eb72f967ed97a4e5e82
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff34d0d9eb9048ebc61b0834b38ecc68abf28009a06896a35319e313d5a3d9e9
+size 161248
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..d14651b521957949d16d7b79f5e97e1c2cf7f85b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit and tie, sitting in front of a microphone, likely participating in an interview or a press conference. The man is bald, wearing a dark suit with a white shirt and a patterned tie. He appears to be speaking or gesturing with his hands, suggesting he is engaged in a conversation or explaining something. The background includes a banner with the words \"DOMINANCE MANAGEMENT\" and \"FINANCE\", indicating that the event may be related to finance or business management. The style of the video is a straightforward, professional interview or press conference, with a focus on the man and his interaction with the microphone and the audience."
+    ],
+    "video_ids": [
+        "VGOqnwMCoJw_29_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, a microphone, and a banner.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man dressed in a suit and tie, speaking into a microphone, with a banner in the background featuring logos and text. All core elements described in the 'Object(s)' condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man in in a suit and tie, speaking into a microphone. which a banner in the background. text and text. The elements elements of in the questionObject(s)' condition are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e83e8e0c5f194ce7336c16740a436748226ccf87
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40b5c02ad6f5441b3cd9b7e68d0b16233fc7d3ab017ff87ce8f969394beaab7e
+size 211776
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..277db446411fd42b12bea3fbd9c60128a727577b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a conversation between two men in a garage. The man on the left, wearing a gray jacket, is gesturing with his hands as he speaks, while the man on the right, dressed in a black jacket, listens attentively. They are standing in front of a blue motorcycle, which is parked on a red stand. The garage is filled with various items, including a shelf stocked with tools and a tire. The scene suggests a casual, friendly discussion, possibly about the motorcycle or other topics of interest. The overall atmosphere is relaxed and comfortable, indicative of a friendly gathering or a shared passion for motorcycles."
+    ],
+    "video_ids": [
+        "_pjbeC_YzUI_24_21to162"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, one in a gray jacket gesturing, the other in a black jacket listening. A blue motorcycle and a red stand.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men: one in a gray jacket who is gesturing while talking, and the other in a black jacket who is listening. A blue motorcycle is prominently visible in the foreground, and a red stand is also present under the motorcycle. All elements described in the condition are present and do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two men in one in a gray jacket who appears gesturing, speaking, and the other in a black jacket who appears listening. There blue motorcycle is visible displayed in the foreground, and a red stand is also present, the motorcycle. The elements in in the condition are present and match not contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..799488bb38c6efce476dbbed36ad3c98a86a0812
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b31b1351abbdc8bb028bc8c247cca744e2aace35af2789e649f4a321dc2f2816
+size 221029
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..27bcdc5f6f4fdbf56e748b669aaa7c39747290d6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a group of people enjoying a boat ride on a sunny day. The boat is filled with passengers, some of whom are standing and others are sitting, all of them looking out at the city skyline in the distance. The boat is moving through the water, leaving a trail of white foam behind it. The passengers are dressed in casual clothes, and some of them are wearing sunglasses to shield their eyes from the bright sunlight. The city skyline is a mix of tall buildings and skyscrapers, with the sun shining brightly on them. The overall style of the video is a casual, candid snapshot of a day out on the water, with the focus on the people and their enjoyment of the cityscape."
+    ],
+    "video_ids": [
+        "_R6dFSO730M_31_0to169"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A boat filled with passengers, a city skyline? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a boat with passengers on board, and a city skyline is visible in the background across the water. These elements match the core description provided, with no contradictions observed.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a boat filled passengers, it, and in city skyline in visible in the background. the water. The elements match the description description provided, indicating no additional or.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6f38ce5e811ba801bf455743af4bb10fa8968139
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6df4dc0616e6eb42c111f6816d3c4bf5d59178a0c98542f2e930aba2390da519
+size 323968
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..ce21ee8ef013c3a162283ffea30a170263300e83
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a soccer match. The main focus is on a player in a blue and red jersey, who is skillfully dribbling the ball down the field. He is in the center of the frame, with his body leaning forward in anticipation. His eyes are focused on the ball, which is slightly ahead of him, indicating his intent to control it.  In the background, two players in yellow jerseys are in pursuit. They are running towards the ball, their bodies leaning forward in an attempt to intercept it. Their positioning suggests they are working together to defend their goal.  The field is a vibrant green, contrasting with the players' colorful jerseys. The perspective of the video is from the side of the field, providing a clear view of the action. The focus is on the players and the ball, with the background slightly blurred, emphasizing the intensity of the moment.  Overall, the video captures the excitement and skill of soccer, with the players' movements and the ball's trajectory creating a sense of anticipation and excitement."
+    ],
+    "video_ids": [
+        "67cH6swQElc_32_107to252"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a blue and red jersey, a player in a yellow jersey (pursuing), another player in a yellow jersey (pursuing), and a soccer ball.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a blue and red jersey (Barcelona kit) dribbling the soccer ball while being pursued by two players in yellow jerseys (Villarreal kit). All core elements described are present, and there are no conflicting elements that contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts two soccer in a blue and red jersey dribthecelona)) dribbling a ball ball, being pursued by a players in yellow jerseys.likelyillarreal kit). The the elements of in present: including the are no additional additional. contradict the description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..df6d3a630ba2df2e2d0eb35056d92b1320be300b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1373-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:faecdfc402400f46e6639ac0bba4b2fc55f3f3d82db0605dbb0f45647b7682b1
+size 313926
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..3dbf12a5de3bf2ed4f47b4b42daa9c651ba7ec28
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a chef in a professional kitchen, preparing a meal. The chef is wearing a white chef's coat and is focused on the task at hand. The kitchen is well-equipped with stainless steel appliances, including a stove and a sink. There are also several pots and pans hanging from the wall, indicating that this is a well-used and functional kitchen. The chef is using a cutting board to chop tomatoes, which are placed on the counter in front of them. The overall style of the video is realistic and informative, likely aimed at teaching viewers how to prepare a specific dish. The chef's actions are clear and precise, demonstrating the proper technique for chopping tomatoes. The kitchen setting is clean and organized, providing a professional backdrop for the cooking process."
+    ],
+    "video_ids": [
+        "LktiA-rMOrc_1_0to166"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A chef in a white chef's coat, a cutting board, tomatoes, and several pots and pans hanging from the wall.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chef in a white chef's coat standing at a counter with a cutting board, handling tomatoes, and there are multiple pots and pans hanging on the wall behind her. All elements described in the condition are present, and there are no contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chef in a white chef's coat, at a counter, a cutting board. chopping what. and several are several pots and pans hanging from the wall behind him. The elements of in the condition are present and and the are no contradictions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e00c75ce224b5bd9e2eed43368fb506dff937483
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f211f1c9a448eae92462e8c17e40ef1e6fe93ef367ebf061771f2e131f87daac
+size 149351
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..b86bc038a53fef6c3017ee3470509c0b307742bc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a dish of cooked shrimp, which are seasoned with red pepper flakes and black pepper. The shrimp are pink and white, and they are arranged in a circular pattern in a glass dish. The dish is placed on a wooden table, and there are green beans in the background. The style of the video is a simple, straightforward food presentation, with a focus on the shrimp dish. The lighting is bright, and the colors are vibrant, highlighting the textures and flavors of the shrimp. The video does not contain any text or narration."
+    ],
+    "video_ids": [
+        "SffWUBxXQ9E_30_0to113"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Cooked shrimp, green beans? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows cooked shrimp and green beans in a glass container, which matches the 'Object(s)' condition. The shrimp are pink and seasoned, and the green beans are vibrant and fresh, confirming the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows cooked shrimp arranged green beans. the dish bowl. which align the descriptionObject(s)' condition. The shrimp are presented and white, and the green beans are fresh green fresh, indicating the presence description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bf847a69d740fd9efef85ba00a7ae1a7a7f27dfe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95ab33185cf1fd42da7b07a46f9c0985d8eb15d58f0e07b72b02e817c283fbcd
+size 116653
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..89ee515240f79f133b45e1039c9b2b7b9ce14fe9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse of a car driving down a winding mountain road. The car is yellow and is seen in three different positions as it moves along the road. The road is surrounded by a rocky cliff on one side and a steep drop-off on the other. The sky is clear and blue, and the mountains in the distance are covered in green trees. The car's movement is smooth and steady, and the overall style of the video is dynamic and adventurous."
+    ],
+    "video_ids": [
+        "5l34OPap1IY_22_18to209"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A yellow car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a yellow car driving along a winding road through a mountainous landscape. The car is the primary focus and is consistently visible throughout the frames, fulfilling the 'Object(s)' condition as described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features shows a yellow car driving on a winding road. a mountainous area. The car is the central object and matches consistent yellow throughout the frames, fulfilling the 'Object(s)' condition of described.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..46041015037b4f926dc192a5cb3efc54ecdf4821
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c7b3b0bd9d53f6476ac846cb362c5ef97507f0d5c8beaa1a1d9fb3d784ad2ad7
+size 497392
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..b421382c966a58262bdffd1c8c89ea26103359df
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene outdoor scene featuring a fluffy cat perched on a wooden deck, intently observing a pond filled with koi fish. The cat, with its brown and white fur, is positioned near the edge of the deck, leaning slightly forward to get a closer look at the water below. The pond is home to several koi fish, displaying vibrant colors such as orange, white, and yellow, which swim gracefully across the surface. The water appears calm, with gentle ripples created by the movement of the fish. The wooden deck, made of planks, provides a stable platform for the cat, while the background includes a concrete wall and a wicker basket, adding to the tranquil setting. The lighting suggests it is daytime, with natural light illuminating the scene, enhancing the vivid colors of the koi fish and the texture of the cat's fur. Throughout the video, the cat remains focused on the pond, occasionally shifting its gaze slightly but maintaining its overall posture of curiosity"
+    ],
+    "video_ids": [
+        "f551f8971a6b8fef2ca3c0a7a4f637eb5cb92903b76ad9aa214677d6b79d50a1"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A fluffy cat and koi fish? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a fluffy cat sitting on a wooden deck near a pond, observing koi fish swimming in the water. The core objects described \u2014 a fluffy cat and koi fish \u2014 are present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a fluffy cat and on a wooden platform, a pond filled observing koi fish swimming in the water. The presence elements, in a fluffy cat and koi fish \u2014 are present and fulfill depicted in the scene.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4c85e937f74a709ec157a7aa6c54d264e5ad3559
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:44ff638506591dd7b60f3b4eeae12ea97228cf0e9356f9a9195980c4b2b9380f
+size 262960
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..bb932a98b01863653c652484a4118082a41859ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a cooking process where pieces of meat are being grilled over an open flame. The meat appears to be chicken, with some pieces showing a golden-brown sear while others remain pale and raw. The grill surface is composed of glowing embers and charred remnants, indicating a high-heat cooking method. Smoke rises from the meat and the embers, adding a dynamic element to the scene. The camera remains stationary throughout the sequence, focusing on the interaction between the meat and the heat source. The lighting highlights the texture of the meat and the intensity of the flames, creating a vivid and appetizing visual. There are no other objects or characters present in the frame, keeping the viewer's attention solely on the cooking process. The overall atmosphere suggests a rustic and traditional cooking technique, emphasizing the natural flavors and textures of the food."
+    ],
+    "video_ids": [
+        "8cdeca71b29636ab2aab536c8ecab194eaf58c012169f6819b862c8ad5eb3323"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Pieces of chicken meat, glowing embers, charred remnants? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows pieces of chicken meat being cooked over glowing embers and charred remnants, which aligns with the described 'Object(s)' condition. The smoke, heat, and visible charring further confirm the authenticity of the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts pieces of chicken meat being grilled on glowing embers, charred remnants, which aligns with the ' 'Object(s)' condition. The presence and the, and the cookingarring further support the cooking of the cooking,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..999d00db904844a8775928cb38a0a85338243be3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:36b0404483ce8c597a3e01b9f8fcdf95d86d19d4d0a25e1f7ba058c84445fcc7
+size 166982
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..14e01243eadae082cda2659e3f4a7c61fe253539
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a pile of colorful clothing items, including shirts, pants, and sweaters, in various patterns and colors. The pile appears to be in motion, with the clothes shifting and moving around each other. The style of the video is dynamic and vibrant, with a focus on the textures and colors of the clothing. The camera angle is low, looking up at the pile of clothes, which adds to the sense of movement and energy in the scene. The video captures the essence of a busy, colorful, and chaotic environment, possibly a thrift store or a clothing donation center."
+    ],
+    "video_ids": [
+        "MFpNYQd2t64_11_0to120"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Shirts, pants, and sweaters.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a pile of clothes that includes various garments such as shirts (e.g., the blue satin-like shirt), pants (e.g., the denim jeans visible at the bottom), and sweaters (e.g., the beige knitted sweater on the left). These items match the 'Object(s)' condition of shirts, pants, and sweaters.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a pile of various, appear various shirts such as shirts,asvid., t colorful shirt shirt shirt and pants (e.g., the red-like), in the bottom), and sweaters (e.g., the red sweateritted sweater). the right). The items collectively the 'Object(s)' condition of shirts, pants, and sweaters.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e4fd26200895904def04662b9101e5c692950cb0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e82f2e0204ce0b4207a56e2dbc5bd79208710dbd6ec55fb43e2e4a703bd055c
+size 192848
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..e1f2cd7adf2ac171b2eb364f3479bf3d10bff25e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man playing a guitar. He is wearing a black t-shirt with white text on it. The guitar is black and has a white pickguard. The man is sitting down and is holding the guitar with both hands. He is playing the guitar with his fingers. The guitar has six strings. The man is looking at the guitar as he plays it. The guitar is the main object in the video. The man is the only person in the video. The video is in color. The man is the only object in the video that is moving. The guitar is the only object in the video that is not moving. The man is the only object in the video that is making a sound. The guitar is the only object in the video that is not making a sound. The man is the only object in the video that is wearing something. The guitar is the only object in the video that is not wearing anything. The man is the only object in the video that is holding something. The guitar is the only object in the video that is not holding anything. The man is the only object in the video that is playing something. The guitar is the only object in the video that is not playing anything. The man is the only object in the video that is sitting down. The guitar is the only object in the video that is not sitting down. The man is the only object in the video that is looking at something. The guitar is the only object in the video"
+    ],
+    "video_ids": [
+        "ICd_TqtUWbA_2_0to152"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ['man', 'guitar']? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man playing a guitar, which fulfills the 'Object(s)' condition. Additional elements like other guitars on the wall do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man playing a guitar. which fulfills the 'Object(s)' condition. The elements such the people or the floor in not contradict the core description and}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..341a992f56f00eb47c4438c39e7935cc818a71b3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a4b413db920e24831638c26594b1ffdc2559dfe4989328c6e3a691cc59f1278
+size 220363
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..3d6e25943b49c8b6927cc731c84518313a37b0f7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a large, modern barn-style building with a gray metal roof and red trim. The building is situated on a grassy field with a well-maintained lawn and a few trees. The first frame shows the building from a distance, highlighting its size and the surrounding landscape. The second frame provides a closer view of the building, showcasing its architectural details and the landscaping around it. The third frame offers an even closer perspective, focusing on the entrance and the landscaping in front of the building. The style of the video is a real estate or architectural showcase, with a focus on the building and its surroundings. The video likely includes a voiceover or text describing the building's features and location."
+    ],
+    "video_ids": [
+        "3fufbdmjo8g_1_0to104"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Large, modern barn-style building, grassy field, well-maintained lawn, trees? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a large, modern barn-style building with a distinctive roof and wooden accents. It is surrounded by a well-maintained lawn and grassy field, with trees visible in the background. These elements align with the description provided, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a large, modern barn-style building with a red red and red siding. The is situated by a well-maintained lawn and ay field, which trees visible in the background. The elements align with the description provided, making there conflicting elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cd8dc1a8b6a4b099787a55098032004181f4d861
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1374-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:961f78c7ec01554262a086fa7387957bccdaa0877693f1eccb00f6a2a32f8178
+size 214098
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..19a23bc72430ac39a8788eb30970abf6c0146244
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a nighttime scene in what appears to be a residential area. The primary focus is on a bright light source, likely a security camera or a flashlight, illuminating a patch of ground covered with small rocks and pebbles. A tree trunk is visible on the left side of the frame, partially obscuring the light. In the background, there is a house with some lights on, indicating it might be occupied.  As the video progresses, a deer enters the frame from the right side, moving slowly across the path. The deer's presence adds a dynamic element to the otherwise static scene. The light from the camera continues to cast a strong glow, creating lens flares and reflections on nearby surfaces, such as a plastic container lying on the ground near the center of the frame. The overall atmosphere remains calm and quiet, with no other significant movement or activity occurring in the scene. The camera remains stationary throughout the video, maintaining a consistent perspective on the illuminated area and the approaching deer."
+    ],
+    "video_ids": [
+        "973e7ed284eac65f0ae1055fcfe5d442072cad17363105d9840dec4db6545654"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bright light source, a tree trunk, a house, a deer, and a plastic container.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a bright light source (creating lens flare), a tree trunk on the left, a house in the background, a deer standing on the right, and a plastic container (visible as a distorted, circular object near the light). All elements described are present, even if the deer and container are somewhat obscured by lens effects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a bright light source,a a flare), a tree trunk ( the left, a house with the background, a deer walking on the right, and a plastic container onwhite on a white white possibly object on the center source The elements match in present and and if some deer is the are not abstract by the flare.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..60cd580301457b8a91dc4c9c3dd7d149925b8155
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:066a51a43f4911d09f0430f7379597359dacc97ad447a57687d6a3efefdaa512
+size 210206
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..e4eca7a60835b4b18ac6429c203d430529dee57e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene by a pond where a fluffy brown cat is intently watching a group of koi fish swimming gracefully in the water. The cat, positioned on a stone ledge, leans forward, its attention fully captivated by the vibrant colors and movements of the fish. The koi display a variety of hues, including white, orange, and yellow, with some featuring intricate patterns on their scales. The water is clear enough to see the fish clearly, and small ripples form around them as they swim. The background reveals a tranquil setting with a mix of greenery and a calm surface of the pond. The camera remains steady throughout the sequence, focusing on the interaction between the cat and the fish, highlighting the peaceful coexistence of the two animals in this natural environment."
+    ],
+    "video_ids": [
+        "8d723d214357a4b0889207b2bd05192d832ed55fcee3e61d6207cfaf71f67f83"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A fluffy brown cat and a group of koi fish.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a fluffy brown cat sitting by the edge of a pond, observing a group of koi fish swimming in the water. The cat's fur, color, and posture are consistent with the description, and the koi fish are visible and active in the water. No elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a fluffy brown cat and near a edge of a pond, observing a group of koi fish swimming in the water. The presence's fur is the, and the, consistent with the description of and the presenceoi fish are visible in swimming in the pond, There additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..11a03c3d7291e47ebb03337840aabbdea2efc5c1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c536a76502e43f810f560d41c0c75e01ad7ea6b782ba2af2a4d0d3e4952901a
+size 224052
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..199fb85833d88ae1114e3c6d74acc3c21617ae85
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man in a beige shirt and a beard, standing on a busy street in India. He is holding a small bowl of food in his hands, seemingly enjoying his meal. The street is bustling with activity, with people walking around and cars driving by. There are also several umbrellas and signs visible in the background, adding to the vibrant atmosphere of the scene. The man's casual attire and the lively surroundings suggest a relaxed and enjoyable day in the city."
+    ],
+    "video_ids": [
+        "TcrfmJQbK00_17_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a beige shirt with a beard, holding a small bowl of food.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a beige shirt and having a beard, holding a small bowl of food. These elements are consistent with the description, and the surrounding environment does not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a beige shirt and holding a beard, holding a small bowl of food. The elements match consistent with the description provided and there additional environment, not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ea6d8423158429ae514e1d845ff35ff19b58c386
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a3efd866237233d2ecf7c8f1346e32dffdd4d344200bb410d8c34095ead87fc2
+size 216917
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..807ed6df163feb32fdf4153ce93bd1ab52626a03
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a group of five people are gathered around a desk in an office setting. The office has a modern design with a glass wall in the background. The group consists of three men and two women. One of the women is seated at the desk, while the others stand around her. They are all engaged in a discussion, with the woman at the desk gesturing as she speaks. On the desk, there is a computer monitor displaying some information. The office is well-lit, with natural light coming in through the glass wall. The overall atmosphere of the video is professional and focused."
+    ],
+    "video_ids": [
+        "5NklULEIjz4_165_25to224"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Five people (three men and two women), a desk, a computer monitor.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows five people (three men and two women) gathered around a desk with a computer monitor. The core elements described are present, and while there are additional elements (like plants and a glass block wall), they do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts five people,three men and two women) gathered around a desk with a computer monitor. The presence elements of in present, and there there are additional elements likesuch a and other window wall wall), they do not conflict the main.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1cce4c438229a68a4e81ab6e36131a27770599a8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea788b66829701b716497e0207b0fe6b363a588bd16a9c65f7713b0082245dcc
+size 131912
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..11bbea5a2af7ab7f7fd9ec2ef324e9efdc6ec62e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and glasses, wearing a pink shirt, sitting in front of a brick wall with a neon sign. He appears to be in a room with a bookshelf filled with books and a colorful poster on the wall. The man is gesturing with his hands, suggesting he is in the middle of a conversation or presentation. The overall style of the video is casual and informal, with a focus on the man and his surroundings."
+    ],
+    "video_ids": [
+        "qWuaU_A84yw_12_23to170"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and glasses, wearing a pink shirt? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and glasses, wearing a pink shirt, which matches the description. The background elements, such as the neon sign and bookshelf, do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a man with a beard and glasses, wearing a pink shirt. which matches the description provided The background includes, such as the neon sign and bookshelf, do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..31352b0104fce4c474b264adacd6f7e3e86ea877
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3580cfebf62e8af8ec4514ab04438dafccb7df8215608aa4e658a5b00a5d2a58
+size 254645
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..c6edee4352f02b5371f53f079ede1c41ba813d0c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing on a rocky shoreline with a body of water in the background. He is wearing a green t-shirt, a black baseball cap with a star on it, and sunglasses. The man appears to be speaking into a microphone, suggesting that he might be conducting an interview or recording a podcast. The setting is outdoors, with a clear blue sky and calm water. The man's attire and the microphone indicate that he might be a reporter or a podcaster. The overall style of the video is casual and informal, with a focus on the man and his surroundings."
+    ],
+    "video_ids": [
+        "H-KimVoc79s_0_0to190"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a green t-shirt, black baseball cap with a star, and sunglasses. Microphone.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a green t-shirt, a black baseball cap with a star, and sunglasses, and there is a microphone visible near his collar. All elements of the description are accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a green t-shirt, a black baseball cap with a star, and sunglasses. which there is a microphone in in him mouth. The the match the description are present represented in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3f9333b1de458f41b78334fefaffb8c86b06a306
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d6bc012d693b3c9bc47609e21d2e9b9ae8d7d648276d6badcf579c97488bf6cb
+size 122279
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..82b2cfc128ff68b76e2764024f3346df5a1d31a6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment on a soccer field. A male soccer player, dressed in a white jersey with orange and black accents, is in motion. He is running towards the right side of the frame, his right arm raised in a saluting gesture. His left arm is extended, pointing towards the crowd in the background. The crowd, a blur of colors, is a mix of spectators, some of whom are wearing orange and black, possibly indicating their support for the player's team. The player's focused expression and the blurred motion of the scene suggest a high-energy moment during a soccer match."
+    ],
+    "video_ids": [
+        "QqCwyBO1ovY_48_0to186"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A male soccer player, the crowd.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a male soccer player in a white jersey with tattoos on his arms, shielding his eyes as if looking into the distance, which matches the description. The background is blurred but suggests a crowd, fulfilling the 'crowd' condition. There are no elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a male soccer player in a white uniform with orange on his arms, running his eyes, if looking at the distance. which is the ' of The background includes filled but appears a crowd of which the 'crowd' condition. The are no additional in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..671b95062f217c91eb0e6a19998a56ee3176d752
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:94a34801ffd174598a74a126a2c758725b6ff71aba178a4c0f317669105c65f8
+size 274353
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..227615d0ec83890bc265ae5798090179f678b87e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a child's play kitchen set with various toy food items and utensils. The kitchen set is colorful and includes a toy stove, oven, and sink. The child is seen interacting with the toy kitchen set, placing the toy food items and utensils into the toy basket. The child is wearing a chef's hat, suggesting that they are pretending to be a chef. The style of the video is playful and child-friendly, with a focus on the child's imaginative play."
+    ],
+    "video_ids": [
+        "dzYJLwSl0RI_56_126to278"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Toy food items, utensils, toy basket, child, chef's hat? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows toy food items (like sliced fruits and vegetables), utensils (a wooden spoon and other cutlery), a toy basket (wooden bowl with handles), a child (represented by a plush toy figure), and a chef's hat (worn by the plush toy). All these elements are clearly visible and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful a food items,like cupcakes vegetables and vegetables), utensils (such toy spoon), spat cookinglery), a toy basket (onen blocks), a), a child,a by a young toy),), and a chef's hat.worn by the child toy). The these elements are present visible and match the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bbed67e29a4863fd36e371f8854334133bb5ed4b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1375-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f207e76a23becc5022976711cd6fbc291868a2f01227bf626497bf5b708ce72c
+size 184415
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..4e5b6815c7e09085a1bc3cbdb42011cdbc3683cf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen enjoying a drink from a coconut. He is wearing a white t-shirt and a backpack. The coconut is being held by another man who is wearing a green t-shirt. The scene takes place in a market, with various fruits and vegetables visible in the background. The man in the white t-shirt is seen taking a sip from the coconut, while the man in the green t-shirt is holding it steady. The overall atmosphere of the video is casual and relaxed, capturing a moment of enjoyment in a bustling market setting."
+    ],
+    "video_ids": [
+        "OCS8KhakJhY_109_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a white t-shirt, a man in a green t-shirt, and a coconut.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a white t-shirt drinking from a coconut with a straw, and another man in a green t-shirt holding the coconut. These core elements match the description exactly, and no contradictory elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two man in a white t-shirt and from a coconut, another straw, and another man in a green t-shirt holding the coconut. The elements elements match the description of, fulfilling there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..df26c15818a9dd9c440fdb2bb455a64befcf01ea
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0ec878bfbe9826bc73d1ae1ad823dd6e6bcbbbcdca0fbb1e41d58cb8653506b
+size 192118
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..2e7d54d3083b80972a748fd579ada78d91831e56
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures an intense moment in a basketball game between the Detroit Pistons and the Cleveland Cavaliers. The player from the Cavaliers, wearing a blue jersey with the number 4, is in possession of the ball and appears to be preparing to make a shot. The player from the Pistons, wearing a white jersey with the number 30, is closely guarding him, trying to block the shot. The background shows a packed basketball arena with spectators watching the game intently. The atmosphere is filled with anticipation as the players and the audience await the outcome of the play."
+    ],
+    "video_ids": [
+        "X36k2CyE0HY_0_0to102"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Cavaliers player in blue jersey with number 4, Pistons player in white jersey with number 30, basketball? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a Cavaliers player in a blue jersey with the number 4, and a Pistons player in a white jersey with the number 30, both actively engaged in a basketball game. The basketball is also clearly visible in the player's hands. All core elements described are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a basketball player in a blue jersey with the number 4 and a a Pistons player in a white jersey with the number 30. both actively engaged in a basketball game. The presence is clearly visible visible in the player's hand, The elements elements of in present and match represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3316951e6402458ce88f86af1e56d67c38cd0f49
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4693e0347bc67cd73e80e78b7a044b3267b6cff498a33becfe832b2b33cdbc2d
+size 414115
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..dfe936009999bac47c8fee8290b3cd618ec06f7f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a playful scene involving a blue plastic tube and a series of chocolate animal figurines placed on a curved surface. Initially, two blue balls are seen rolling down from the top left corner of the frame into the tube. As they enter the tube, one ball exits at the bottom right, while the other continues its journey through the tube. The camera remains stationary throughout the sequence, focusing on the tube and the surface where the chocolate figurines are positioned.  As the balls exit the tube, they roll onto the surface, which is lined with various chocolate animal figurines, including a rabbit, a bear, and others. The balls continue to roll past these figurines, creating a dynamic interaction between the moving objects and the static display. The background features a simple indoor setting with a white wall and some wooden blocks supporting the curved surface, adding a sense of depth and context to the scene. The overall atmosphere is light-hearted and engaging, capturing the viewer's attention with the movement and the charming"
+    ],
+    "video_ids": [
+        "6952e4ce9128484b134213da955256b1346223e8d2e7f25489e9ba8806492f07"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two blue balls, a blue plastic tube, chocolate animal figurines (rabbit, bear, others)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows two blue balls rolling down a blue plastic tube and along a track. Chocolate animal figurines, including a rabbit and a bear, are visible in the background, matching the description. The scene is consistent with the specified objects without contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful two blue balls being out a blue plastic tube, interacting a surface, The animal figurines, including rabbits rabbit and a bear, are present in the background, fulfilling the description. The presence does set with the objects objects and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..291e5640cc4f09f6ceb27b917e19c55ea9ef3461
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9d880404e62607ae35f48d082d436568542ea9d07553ce3e01d4a02d044d5542
+size 167842
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..5e416fc29949dbe08690b43111882848959a5223
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person in a red apron preparing a meal in a kitchen. The person is using a silver pot on a gas stove, and a black and silver pressure cooker is nearby. The pressure cooker has a digital display showing the time as 03:45. The person is pouring something from the pot into the pressure cooker. The kitchen has a modern design with a white countertop and a black stove. The person is focused on the task at hand, and the overall atmosphere of the video is one of careful meal preparation."
+    ],
+    "video_ids": [
+        "Yd-SNqOf0-U_11_0to102"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Person in a red apron, silver pot, gas stove, black and silver pressure cooker, white countertop, black stove.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition. It clearly shows a person wearing a red apron, a silver pot being placed on a gas stove, a black and silver pressure cooker (with visible brand 'Instant Pot' and digital display), and a white countertop beneath the stove. The stove itself is black, matching the description. All core elements are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as It shows shows a person wearing a red apron, a silver pot, lifted on a gas stove, and black and silver pressure cooker,which a control markingsT Pot'), on ' display), and a white countertop in the stove. The black is is black with and the description. The elements elements of present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..18d2c37bf7ae8df988e3b4d5b9de5152908a16ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d6a43dc79b85b595283ed026199cf81c2f3418072ca246fd83626c347b2e09c7
+size 201218
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..b08d258648172819a7d5d1462f5d71d2334387a8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video opens with a serene view of a lush, verdant forest, where dense green foliage dominates the landscape. A waterfall cascades down a rocky cliff in the background, adding a dynamic element to the tranquil setting. The camera remains stationary throughout the sequence, allowing viewers to take in the natural beauty of the scene. In the foreground, a river flows gently, reflecting the surrounding trees and sky. A bird, possibly a heron or egret, is seen flying gracefully over the water, its wings spread wide against the backdrop of the forest and waterfall. The bird's movement is smooth and fluid, creating a sense of harmony between the animal and its environment. The overall atmosphere is peaceful and untouched, capturing the essence of nature in its raw form."
+    ],
+    "video_ids": [
+        "8ae68cbf5c4662cd2d5d3898ab00efc6a52e1d0e0782fe8ee6587f1f2a33fe45"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bird, possibly a heron or egret, flying gracefully over the water.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bird, consistent with the appearance of a heron or egret, flying gracefully over a body of water. The bird's flight path and the surrounding lush greenery and water align with the description. The presence of trees and a fallen log does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts a serene, which with the description of a heron or egret, flying gracefully over the body of water. The bird is presence path is the context environment greenery align waterfallfalls with the description of The presence of the and the waterfall log in not contradict the core description of}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..831fb0506f1e74071bfe074b5636ba015c248594
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ca69c39670fc99eb98da7c49d2e72b3536951d96028d5c3d9a8e5ea6dbd4b8a7
+size 206753
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..4b65887322478f1d4a92f756e5a3a6388eab49a8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen in a kitchen, enjoying a drink from a mason jar. She is standing in front of a counter that is adorned with various items, including a blender, a vase of flowers, and a potted plant. The kitchen is well-lit, with a window providing natural light. The woman is dressed casually, wearing a black shirt and a necklace. The overall atmosphere of the video is relaxed and comfortable, capturing a simple yet enjoyable moment in the woman's day."
+    ],
+    "video_ids": [
+        "UvC7X3Gchtk_9_121to265"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a mason jar, a blender, a vase of flowers, and a potted plant.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman holding a mason jar, with a blender visible on the counter, a vase of white flowers on the left, and a watermelon (which can be considered a potted plant or fruit in a container) on the counter. All specified objects are present and correctly identified.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a woman in a cupason jar, which a blender and on the counter. a vase of flowers flowers, the left, and a pmelon plantwhich can be considered a potted plant) a) a pot) on the right. The the objects are present and match identified.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8a1dff525f2f0f61c45bd71d1231cbf56a327f8a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b00ada0fc74fe9581303c13634bc767472bcb7560b6c815fbba874072e5be8cf
+size 160043
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..488765ae7b054b394fa9263b1f40686d7c6fc000
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is a player wearing a maroon jersey with gold accents, the number 8 prominently displayed on his chest. He is in a defensive stance, crouched low to the ground, ready to spring into action. His helmet is gold with a maroon logo, matching his jersey.  In the background, there are other players, some in the same maroon and gold uniforms, while others are in white jerseys. The field is a vibrant green, contrasting with the players' colorful attire. The crowd in the stands is a blur of colors, indicating a large audience for the game.  The style of the video is a fast-paced, action-packed sports sequence. The camera angle is low, emphasizing the player's readiness and the intensity of the game. The focus is sharp on the player in the maroon jersey, while the background is slightly blurred, drawing attention to the main action. The lighting is bright, suggesting an outdoor setting during the day. The overall impression is one of anticipation and excitement, typical of a live football game."
+    ],
+    "video_ids": [
+        "4706v5fIUes_15_128to301"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a maroon jersey with gold accents, number 8, and another player in a white jersey.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a player in a maroon jersey with gold accents and the number 8, as well as another player in a white jersey. The description matches the visual content without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a player in a maroon jersey with gold accents and the number 8. which well as another player in a white jersey. The player matches the core elements of any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d2dd648f66f3062a47534530889397a6c226ba93
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5c1a701cacd1c29758f6f51fc23dd08ff55cbd43e4803d696c2ef548f6f0dde
+size 180673
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..8851d34787690263044f750ed40beda32705579f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a vintage car driving down a snowy road. The car, a silver Mercedes-Benz, is seen from the front, with its headlights on, illuminating the snow-covered road ahead. The car's license plate reads \"S-MB 280H\". The road is flanked by snow-covered fields on both sides, and power lines run parallel to the road. The sky above is a clear blue, suggesting a cold winter day. The car's speed is indicated by the motion blur, giving a sense of movement and travel. The overall style of the video is a blend of vintage and modern, with the classic car juxtaposed against the contemporary snowy landscape."
+    ],
+    "video_ids": [
+        "l0MbabZOEbs_23_0to111"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A silver Mercedes-Benz car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a silver Mercedes-Benz car driving on a road, matching the core description. The car's distinctive grille, headlights, and license plate are visible, confirming it is a Mercedes-Benz. The presence of snow and other background elements does not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a silver Mercedes-Benz car driving on a snowy covered which the description description of The car's design shape and headlights, and overall plate are visible, confirming it is a Mercedes-Benz. The presence of the and a elements elements does not contradict the main of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6063b0805d7f7c1678aeaf3cb2796735cd3823da
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1376-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd2b306227ef2851484bd7e2397a917d7a3af3b219d1c32f8144d86b195820e4
+size 323812
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..963485562cef6a64c8dc6fc58133709015b69e02
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a classic car parked on a gravel road. The car is black with a silver grille and bumper. The license plate reads \"B37-215\". The car is facing the camera. The background features a grassy field and trees. The car appears to be in good condition. The style of the video is a straightforward, clear shot of the car without any additional action or movement."
+    ],
+    "video_ids": [
+        "u9CMrLbY5S8_88_72to199"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black car with a silver grille and bumper, license plate 'B37-215'.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black car with a silver grille and bumper, and the license plate clearly reads 'B37-215'. The description matches the visual content accurately, with no contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a black car with a silver grille and bumper, and the license plate ' reads 'B37-215'. The car matches the video elements of, with no additional or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9aaa6b0baec214ca670b4a1db7fac632226b1e2f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fa4e5a6e32075b4978752ce3c407dd4b9a28eed211e67dbda18bd1d1cfb35ad7
+size 82814
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..4689359bbca0278ac9524e91a2fadfe2291c55bc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a wooden desk surface, showcasing various objects arranged on it. The primary focus is on a black computer mouse positioned near the center-left of the frame. Adjacent to the mouse, there's a small bottle of glue labeled \"Wilt You Quit,\" suggesting a crafting or DIY activity might be underway. A piece of paper lies beneath these items, partially covered by a red plastic track, which resembles a toy train set. The track curves gracefully across the paper, adding a playful element to the scene.  As the video progresses, the camera remains stationary, maintaining a consistent angle and distance from the objects. The lighting is steady, highlighting the details of each item without any noticeable changes in brightness or shadows. The background remains unobtrusive, ensuring that the viewer's attention stays on the desk and its contents.  Throughout the video, there are no visible characters or animals present, and the scene remains static except for the slight movement of the camera itself, which subtly"
+    ],
+    "video_ids": [
+        "acbcc1337ba99a8190604052bc2b98d210812f0342fdc46e50817bc726d57d5f"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Black computer mouse, small bottle of glue labeled 'Wilt You Quit', piece of paper, red plastic track resembling a toy train set.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a black computer mouse, a small bottle of glue labeled 'Wilt You Quit' (which appears to be a playful misspelling of 'Wite-Out'), a piece of paper, and a red plastic track resembling a toy train set. All these items are visible and consistent with the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showing showing a black computer mouse, a small bottle of glue labeled 'Wilt You Quit', (which is to be a play orpelling of 'Willilt-Out'), a piece of paper, and a red plastic track resembling a toy train set. The these objects are present and match with the description provided.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cac2b1967c155065b61ba76be97b93260a2b3830
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:80b7bd00607ac0c9027851d4de037608d216a2275ac90f7a56cbaa3b099b3475
+size 98473
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..0993432b7cf7d60d265d54400dba1970b729d179
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with red hair, wearing a black and white striped shirt. She is standing in front of a blue background with a graphic of a globe. The woman is looking directly at the camera, and her expression is neutral. The style of the video is a news segment, as indicated by the logo in the corner of the screen. The woman appears to be a news reporter or anchor, and the background suggests that the segment is about global news or events. The overall tone of the video is professional and informative."
+    ],
+    "video_ids": [
+        "XtFBWPOSz38_21_0to110"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with red hair, wearing a black and white striped shirt, and a news segment logo in the corner of the screen.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with red hair wearing a black and white striped shirt, and there is a news segment logo (with 'D NEWS' and a globe graphic) visible in the top right corner of the screen. These elements match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with red hair wearing a black and white striped shirt. which there is a news segment logo inG theGIF' and a globe design) in in the corner left corner of the screen. The elements match the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..92bb208360e2e16e00beb7747410ff51f79198d1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3a4215ef0a3626457e7613b37d6253097b25a9fe4ac50236021e9080899fe71
+size 107557
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..31b89632cf8d70eb2b117b3aa99339cd7bec4394
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is standing in a store filled with motorcycle gear. He is wearing a purple t-shirt and blue jeans. The store has a variety of helmets and jackets on display. The man appears to be speaking, possibly explaining the features of the gear or making a purchase decision. The store has a modern and stylish design, with a clean and organized layout. The lighting is bright and inviting, highlighting the products on display. The man's position in the store suggests that he is the main subject of the video, with the motorcycle gear serving as the backdrop for his actions."
+    ],
+    "video_ids": [
+        "QFhIZa8pJjo_6_44to174"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, various helmets and jackets on display.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man standing and speaking, and in the background, there are shelves displaying various helmets and jackets, which matches the 'Object(s)' condition. The presence of additional elements like the mannequins and the motorcycle does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man standing in gest in with there the background, there are various displaying various helmets. jackets. which align the descriptionObject(s)' condition. The presence of the elements like the shelves'squinins and the store in not contradict the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c728bd7dc8e31569872ee17aacb410b0683a2799
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2a493415b6650845e0483b64d172351785ebc7448c2a890b317e36aa25d146ee
+size 231892
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..b92505e18f2cd83d03eee88064295b58dc80271b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man sitting in front of a blue and white background. He is wearing a striped shirt and appears to be in a relaxed state, with his eyes closed and mouth slightly open. The lighting in the room is soft, casting a gentle glow on the man's face. The overall style of the video is calm and serene, with a focus on the man's peaceful demeanor."
+    ],
+    "video_ids": [
+        "P7JkVsIt1fg_39_0to126"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man sitting, wearing a striped shirt, eyes closed, mouth slightly open? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting and wearing a striped shirt, which matches the description. He is seen with his eyes closed and mouth slightly open at certain moments, fulfilling the specified condition. Additional background elements (neon lights) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man sitting, wearing a striped shirt. with matches the description. His has also with his eyes closed and his slightly open, different points, which the ' conditions. The elements elements likelikeon lights and are not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1492cc86b4ebe105f8e960361a84edcc1a998661
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5858c0bf2493d1715d81555861ba73c6edfd5512f922336f522d69fd72b8c5ee
+size 145018
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..e076786e84605af4f7a299f8ccb59698071543a1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young girl is seen in a playful and colorful kitchen setting. She is dressed in a vibrant orange shirt and a black cape, adding a touch of whimsy to the scene. The girl is standing in front of a table, which is adorned with a variety of objects including a vase, a pumpkin, and a potted plant. The table is set against a backdrop of a wall decorated with a sign that reads \"YUM\", further emphasizing the culinary theme of the video. The girl's facial expression suggests that she is making a funny face, possibly in response to something she sees or hears. The overall style of the video is light-hearted and fun, with a focus on the girl's playful interaction with her surroundings."
+    ],
+    "video_ids": [
+        "IhTJOgRAFlg_0_0to164"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Young girl, vibrant orange shirt, black cape, table, vase, pumpkin, potted plant, girl's funny face? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young girl wearing a vibrant orange shirt with a candy design and a black cape with red trim. She is seated at a table, and the background includes a vase with flowers, a pumpkin, and a potted plant. The girl is also making a funny face, which matches the description. All core elements are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young girl wearing a vibrant orange shirt and a black corn and a black cape. a lining. She is standing at a table with which there table includes a vase, a, a pumpkin, and a potted plant. The girl's making seen a funny face, which is the description. The elements elements of present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ebba32e64b2d8d1af18666105c236f305e20afb6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c7b87fa454ce9f43ace5c16e6d0a613e69a5c54074b9ad2d5ee0b547e5bcf43b
+size 178843
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..7db604a027a14dce533c9c67f5e622abc3a2ad56
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene by a calm body of water, likely a pond or a small lake, surrounded by lush greenery. Two ducks are prominently featured in the center of the frame, standing on a submerged log. The duck on the left is a male mallard, identifiable by its vibrant green head and chest, while the duck on the right appears to be a female mallard, distinguished by her more subdued brown plumage. Both ducks are facing towards the right side of the frame, seemingly engaged in some form of interaction or communication.  The background is filled with dense foliage, including various types of trees and bushes, creating a natural and tranquil setting. The water reflects the surrounding greenery, adding depth and a sense of stillness to the scene. There are no significant changes or movements throughout the sequence of frames; the ducks remain stationary, and the environment stays consistent, emphasizing the peacefulness of the moment captured. The lighting suggests it is daytime, with sunlight filtering through the leaves,"
+    ],
+    "video_ids": [
+        "0233530e4ee0c37c5c354489385f8c774707d48fb5249ab60394334b820c9899"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two ducks (a male mallard and a female mallard), a submerged log? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two ducks, one male mallard (with green head and white neck ring) and one female mallard (with mottled brown plumage), standing at the water's edge. A submerged log is visible beneath the water surface near the ducks, partially obscured by the water's reflection and vegetation. The core elements described are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows two ducks, which male mallard andd a head and blue body)) and one female mallard (with brownottled brown andage), standing on the edge's edge. There submerged log is not in the water surface in the ducks, fulfilling obscured by the r's reflection. the. The presence elements of in present, contradiction.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fbe911b4ee7b132b2dc42030bc6631629761b282
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3508bdf303dc521e7cb1f801ca30c33c66730c04e3f139e43cfae77e4631a660
+size 171730
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..23b2a80c2fc862714dad959a0d8e8edd5021d1fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a kitchen, wearing a blue apron and glasses. He is standing in front of a refrigerator, which is silver in color. The kitchen is well-equipped with various appliances, including a blender and a toaster. The man appears to be engaged in a conversation, as he is looking to his left. The overall style of the video is casual and informal, capturing a moment in the man's daily life."
+    ],
+    "video_ids": [
+        "L5mL7VFtvvo_12_0to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue apron and glasses, a refrigerator, a blender, and a toaster.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue chef's jacket (which can be considered a type of apron), glasses, and is in a kitchen setting. A refrigerator is visible in the background, along with a stand mixer (which can be mistaken for a blender) and a toaster oven. These elements largely match the description, and any minor discrepancies (like the stand mixer instead of a blender) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a man wearing a blue apr's apr andwhich can be considered a blue of apron) glasses, and standing standing a kitchen setting. There refrigerator is visible in the background. and with a white mixer,which can be mistaken for a blender in and a toaster.. The elements collectively match the description provided fulfilling the additional discrepancies (like the stand mixer being of a blender) do not significantly the core description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7266c41a7f0ea39ff73425a71a36029f7cf22802
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1377-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7352c30431c32a90ee280b78a815f20e9f011ed8dfeea701a1c07e0f8a3aee2a
+size 171517
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..b852dd9f931045cbedc6f325654852df9622714e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a vibrant blue betta fish swimming gracefully inside a clear wine glass filled with water. The fish's flowing fins and tail move fluidly as it navigates the confined space. The background is a plain white surface, which contrasts sharply with the deep blue color of the fish, making it stand out prominently. The lighting is bright, highlighting the fish's iridescent scales and the clarity of the water. The fish appears to be exploring its surroundings, occasionally pausing to adjust its position or swim in different directions. The overall scene is serene and visually striking, focusing entirely on the graceful movements of the betta fish within the glass. There are no significant changes in the environment or the actions of the fish throughout the video; it remains a consistent and tranquil depiction of the fish in its aquatic habitat."
+    ],
+    "video_ids": [
+        "6e2a6ba0e078410d5c7d2a0e2e413dd66292586fae1920c2305cdc26ce08ecba"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A vibrant blue betta fish? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a vibrant blue betta fish swimming inside a wine glass filled with water, which matches the core description. The fish's color and species are accurately represented, and no elements contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a vibrant blue betta fish swimming in a clear glass, with water. which align the description description of The fish's color and the are clearly represented, and the additional contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2e63e9a80d38e160e87a439da7c973651aa3f892
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be0573f7ef8b242e5c6af82e911f80d04046ece97e8f482bd1c383cc5e24dccd
+size 136493
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..fa4a4dba67041ad463996ff2cfa1a694a8b448c6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a street scene featuring a man standing in the foreground. He is wearing a gray jacket and a blue shirt. The man appears to be looking off to the side, possibly at something or someone not visible in the frame. In the background, there are parked cars, including a yellow taxi, and a building with a red awning. The street is lined with potted plants, adding a touch of greenery to the urban setting. The overall style of the video is casual and candid, capturing a moment in the man's day as he stands on the street."
+    ],
+    "video_ids": [
+        "5r3i8njkOOs_6_0to145"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a gray jacket and blue shirt, standing in the foreground and looking off to the side.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a gray jacket and blue shirt, standing in the foreground. He is looking off to the side at various points in the video, which matches the description. The presence of background elements like cars and buildings does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a gray jacket and a shirt, standing in the foreground and He is looking off to the side, a points, the video, which align the description. The background of a elements like a and a does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..430e26a5e4e52d685e3f7932ecb6af83b63bdae9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cba3d80750d8701e0af27b8e701dfafd5d14d6ada15d06924e6d911e1efc406d
+size 170186
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..e89715e20b7933097bf2f7fb786c951bc5a3a587
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a black shirt with a white collar, sitting in front of a blue background. He appears to be speaking or reacting to something, as his mouth is open and his eyes are closed. The man is the main subject of the video, and there are no other objects or people visible. The style of the video is a close-up shot, focusing on the man's face and upper body. The lighting is bright, highlighting the man's features and the blue background. The video does not contain any text or additional graphics. The overall impression is that the man is engaged in a conversation or reacting to a situation, but the specific content of his speech or the context of the situation is not clear from the image alone."
+    ],
+    "video_ids": [
+        "KN31FtPgcfE_4_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a black shirt with a white collar? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black shirt with a visible white collar underneath. This matches the description exactly, and no conflicting elements are present that would contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black shirt with a white white collar.. The matches the description of, indicating there additional elements are present.\"\n would contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9004816fba2a2109b5d06e945f7c1375bbcae0e9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:faba4fea31438383df83535fe17c8f31f19f05d5dc391d48ab6d931a322f24e7
+size 143618
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..c12478c718b240e828e882a4e1b39e230da20731
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a young girl with pink and purple hair, sleeping peacefully in her bed. She is wearing a pink and white blanket with a floral pattern. In her arms, she holds a pink teddy bear, which is also wearing a matching pink and white blanket. The bed she is sleeping on is pink and has a floral pattern on the headboard. The room is decorated with a blue and yellow striped wall, and there is a small green plant on the windowsill. The overall style of the video is bright and cheerful, with a focus on the girl's peaceful sleep and her adorable teddy bear."
+    ],
+    "video_ids": [
+        "2bYF10U-q3c_347_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young girl with pink and purple hair, a pink and white blanket, a pink teddy bear, and a small green plant.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young girl with pink and purple hair sleeping under a pink and white blanket, holding a pink teddy bear. A small green plant is visible in the bottom left corner of the frame. All core elements from the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a young girl with pink and purple hair, on a pink and white blanket. holding a pink teddy bear. There small green plant is visible in the background left corner of the frame, The elements elements of the description are present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7018fdf537e611eed74190effe8cb3ecaa3deed5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:344a54154421da813bd79e4d81752f4d08ed9e290197359fb4a431365a6a5c6f
+size 203852
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..c6e85ca3e1c8eb6654d0b5dd28bd8cc6a880183f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a dynamic and stylish advertisement for a red Subaru car. The car is shown in motion, driving down a winding mountain road. The car's sleek design and vibrant red color stand out against the natural backdrop of the road, which is surrounded by lush green trees and towering rocky cliffs. The car's powerful performance is emphasized by the blurred motion of the background, suggesting speed and agility. The overall style of the video is sleek and modern, with a focus on the car's design and performance. The video is likely intended to appeal to car enthusiasts and potential buyers, showcasing the car's capabilities and design in a visually appealing way."
+    ],
+    "video_ids": [
+        "v00UbBY6_mA_25_83to208"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red Subaru car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a red Subaru car driving on a road, which matches the 'Object(s)' condition. The car is the central focus and is accurately depicted as a red Subaru, with no contradictions in its appearance or behavior.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a red Subaru car, on a road. which ful the descriptionObject(s)' condition. The car's the central focus of is consistent depicted as a red Subaru, with no conflicting to the appearance or presence.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..39ef0070edb43519051e22faa3728ecd8772b29e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a4a0ed6ac91e11d15165726e895fc2f6f016d7abe88f695f9c6e5e949bd89505
+size 273080
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..104fefab68f6a6c2533513043bb25e8c14f06fcf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman with blonde hair and a silver dress is seen speaking into a microphone. She is wearing a large pearl necklace and earrings, and her makeup is done in a dramatic style. The setting appears to be a store or boutique, with other women in the background. The lighting is bright and artificial, and the overall style of the video is polished and professional."
+    ],
+    "video_ids": [
+        "3auygxpwIuQ_6_40to168"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with blonde hair and a silver dress, wearing a large pearl necklace and earrings, speaking into a microphone with dramatic makeup.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair and a silver dress, wearing a large pearl necklace and earrings, speaking into a microphone. Her makeup is dramatic, with red lipstick and dark eye makeup. The description is largely accurate, and additional elements in the background do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a woman with blonde hair wearing a silver dress, wearing a large pearl necklace and earrings, speaking into a microphone. The makeup is dramatic, which dark lipstick and dark ey makeup. The setting matches largely matched, and the elements such the background do not conflict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1e88727cf508b3d77f01f8602a9b02c3c9f9917a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d6de22eb63aa5a48a4987a970168dd9ca651f66ca2579c82c9ed3ccbc034e95f
+size 166728
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..c16f483cc81b44f44924975bcd3ed6a05dd6ae3b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen sitting on a blue couch with a large collection of reborn dolls. She is holding one of the dolls in her arms, and the rest of the dolls are arranged around her on the couch. The woman is wearing glasses and a plaid dress. The setting appears to be a television studio, as indicated by the presence of a coffee table and a potted plant in the background. The video is likely a segment from a television show, as suggested by the text overlay that reads \"My 12 children are reborn dolls\" and the hashtag \"#ThisMorning\". The style of the video is a mix of documentary and interview, with the woman being the main subject and the reborn dolls being the focus of the discussion."
+    ],
+    "video_ids": [
+        "RZ5lfsoFsvQ_1_98to267"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a blue couch, a large collection of reborn dolls, a coffee table, and a potted plant.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a woman sitting on a blue couch surrounded by a large collection of reborn dolls. A coffee table is visible in front of the couch, and a potted plant is placed on it. The scene is set in what appears to be a television studio, consistent with the context of the video being from 'This Morning'. All specified objects are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a woman sitting on a blue couch, by a large collection of reborn dolls. There p table is also in the of the couch, and a potted plant is present on the. The presence is consistent in a appears to be a cozy studio, as with the presence of the video being a aOm Morning' The elements elements are present and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..09dd4d5692c66f8eef5f9be7e52005e405d7c998
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9defbc98d981c41f5faa1d258817b44dc1629f05406f97854d2ad932b45972b
+size 234134
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..197cfad1887cadee7e686c8a4834d2c79c5a1fd6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is on two players from the same team, both dressed in white and blue uniforms with the number 3 prominently displayed on the back of one player's jersey. The player with the number 3 is in the process of throwing a pass, his arm extended and the football visible in his hand. The other player, wearing a helmet with the number 6, is on the ground, possibly having just made a tackle or fallen during the play. The background is filled with other players, some in motion and others standing still, all dressed in similar uniforms, indicating that they are part of the same team. The style of the video is a real-time capture of the action, with no visible filters or artistic effects, providing a clear and unobstructed view of the game in progress."
+    ],
+    "video_ids": [
+        "qvjjYjIf1b8_17_0to155"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two players (one with number 3, the other with number 6), other players? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two players: one kneeling with number 6 on his jersey and another standing with number 3. The standing player is clearly visible, and although other players are blurred in the background, they do not contradict the core description. The focus remains on the two specified players, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a players, one wearing with the 3 and their jersey and another standing with number 3 on There standing player is holding interacting, and the the players are present in the background, they do not contradict the core description. The presence is on the two players players, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fae4a36fcb2a9d510f4644279f45f1531db872a4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1378-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7d4b3c330c4dd3a51da570e22ae2649a6246b1c60fad8eb997933001a8a6d7af
+size 194495
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..6b9fd4889f8985301958a068d79bf8ab1a20c171
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a white Ferrari sports car parked on a street. The car is sleek and shiny, with a distinctive yellow Ferrari logo on the side. The car's design is modern and aerodynamic, with large black wheels and a low profile. The car is parked in front of a brick building, which adds a contrasting texture to the scene. The car is the main focus of the video, and its position in the frame suggests it is stationary. The video style is realistic, with natural lighting and colors that accurately represent the scene. The car's position and the surrounding environment suggest that the video was taken during the day. The car's shiny exterior reflects the light, indicating that the weather is clear and sunny. The brick building in the background adds a sense of depth to the scene, suggesting that the car is parked on a street rather than in a garage or parking lot. The video does not contain any text or additional objects, and the focus is solely on the Ferrari sports car."
+    ],
+    "video_ids": [
+        "Kuxxw41aRmk_52_0to173"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white Ferrari sports car with a distinctive yellow logo, large black wheels, and a modern, aerodynamic design.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white Ferrari sports car with a distinctive yellow prancing horse logo on the side, large black wheels with red brake calipers, and a modern, aerodynamic design. These features align with the description provided, confirming the video successfully fulfills the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a white sports sports car with a modern yellow logoancing horse logo, the front, large black wheels, yellow accents calipers, and a modern, aerodynamic design. The features match with the description provided, making that video largely fulfills the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f5ecd3586398e858e03f2cc69294df68326a2d26
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:80cfae495adbcc6f59af52fcfe8dac8a0566e051d38913e4b31755c54286b8d6
+size 182928
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..8cd8a25bd38beb82c0be7f2cf50837a2ac42ebe7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young boy is seen enjoying a snack while sitting in a blue stroller. He is wearing sunglasses and a white t-shirt with black text. The stroller is positioned on a sidewalk, and the boy is holding a bag of popcorn in his hand. The scene is set against a backdrop of a sunny day with trees casting shadows on the ground. The boy appears to be in a relaxed and happy mood, enjoying his snack and the outdoor environment."
+    ],
+    "video_ids": [
+        "5DImaGQEjkQ_28_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young boy, a blue stroller, a bag of popcorn, sunglasses, a white t-shirt with black text.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young boy sitting in a blue stroller, wearing sunglasses, and holding a bag of popcorn. He is also wearing a white t-shirt with black text. All the specified objects are present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young boy sitting in a blue stroller, eating sunglasses, and eating a bag of popcorn. He is wearing wearing a white t-shirt with black text. The the elements elements are present in match the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0df1eee20750cf2fc63b08fa99d495dab7758b8c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:05a839817aa8148324bb4f444d5a246d38970164fdd3b46bba625ddeff3efa94
+size 181978
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..9b244bd547ba09b737bdff8a116267837045a022
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news report featuring a woman standing in a garage. She is holding a microphone and appears to be reporting on a story. The garage is filled with various tools and equipment, suggesting it might be a workspace or a repair shop. The woman is dressed in a stylish orange jacket and has her hair styled in a high ponytail. The news report is likely discussing a recent event or incident that occurred in downtown Los Angeles. The style of the video is informative and professional, with a focus on delivering news to the viewers."
+    ],
+    "video_ids": [
+        "E_zLh-QnPU0_2_85to266"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a microphone? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman holding a microphone, which matches the 'Object(s)' condition. The setting and other elements in the background do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a woman holding a microphone, which ful the 'Object(s)' condition. The woman appears the elements in the background do not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7f1078bdeae0eb6cd888a9b6b333ebee4a8a551a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d55c87b0dfeecfa967a886f53dd72067aa82a9ad956df6c23f0336e4713a11b8
+size 90169
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..24d8a40978000da1ef201007750c285ee3289d8b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a black jacket and a black beanie working in a meat processing facility. He is standing at a metal table, using a machine to process meat. The machine is large and industrial, with a conveyor belt that moves the meat along. The man is focused on his work, carefully handling the meat as it moves along the conveyor belt. The facility is clean and well-lit, with metal walls and floors that reflect the bright lights. The man is the only person visible in the video, and he is the main subject of the video. The video is a straightforward depiction of a worker in a meat processing facility, with no additional elements or distractions."
+    ],
+    "video_ids": [
+        "4QOwAUp2rhI_84_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black jacket and beanie, a large industrial machine with a conveyor belt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black jacket and beanie operating a large industrial meat saw, which matches the described object(s). The machine is indeed an industrial saw, and while it does not have a conveyor belt, the core elements described are present and correctly depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black jacket and aanie, a large industrial machine processing, which is the description ' of). The machine has indeed a industrial machine with and the the is not have a conveyor belt, the presence elements of in present. accurately depicted.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4907a76b07372ce9bfd223482263cc944bcca093
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e279caf77496fcdfb0978dc38d9bfe7ac58e1c79a4fef512f9df1eaed990a901
+size 213794
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..b024640c9c3ab831fa0d056485c00485f5b56eca
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a long beard and curly hair, dressed in a white t-shirt with a number one on it. He is standing against a white background. The man appears to be in a thoughtful or contemplative state, as he gazes off to the side. The lighting in the video is soft and even, highlighting the man's features without creating harsh shadows. The overall style of the video is simple and straightforward, focusing on the man and his expression rather than any additional elements or distractions."
+    ],
+    "video_ids": [
+        "1mEBDZ4ZXnc_44_107to261"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man with long beard and curly hair, wearing a white t-shirt with a number one? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with long, curly hair and a full beard, wearing a white t-shirt with the number '1' clearly visible on it. These elements match the description exactly, and there are no conflicting elements in the frame.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with long, curly hair and a long beard, wearing a white t-shirt with the number '1' printed visible. it. The elements match the description provided, indicating there are no additional elements present the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..722e98b23226752e348f7a848406cdbdff728282
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2246c91c0d8116a4d26f732fba48ac0e382b72a15cee1ecbd361349f1b8ea663
+size 98969
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..cf3639d4c12ba497c966b975e156fd94658b2a95
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a thrilling ride on a motorcycle. The rider, clad in a black leather jacket, is seen maneuvering the motorcycle down a winding road. The motorcycle, painted in a sleek black and white color scheme, is the main focus of the video. The rider's hands are firmly on the handlebars, indicating the need for control and precision.  The road itself is a spectacle, with a white line running down its center, guiding the rider's path. The surrounding landscape is dotted with cars, adding to the sense of speed and movement. The perspective of the video is particularly striking, taken from the rider's point of view, it immerses the viewer in the experience of the ride.  The video is a dynamic blend of motion and stillness, with the motorcycle and rider in constant motion against the static backdrop of the road and cars. The overall style of the video is realistic, capturing the essence of a motorcycle ride with precision and detail."
+    ],
+    "video_ids": [
+        "3WnMJtG4uJ4_12_0to181"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Motorcycle, rider in a black leather jacket? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a motorcycle with a rider wearing a black leather jacket, matching the specified 'Object(s)' condition. The perspective is from behind the rider, and the motorcycle and rider are the central focus, with no contradictions to the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows shows a motorcycle, a rider, a black leather jacket. which the description 'Object(s)' condition. The motorcycle is from the the motorcycle, focusing the motorcycle is rider are the central focus of fulfilling no conflicting to the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2259c6a79b6554910fc40b9559b8950d2ef0c8ec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:05e25ebe938b939bcaba9f999259ce3a4eb91303eead441b075f7ee74db6bc7d
+size 145732
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..787e523994c3069646749a838ab8bd0e4499b25b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment from MSNBC, featuring two politicians, a man and a woman, walking down a hallway. The man is wearing a suit and tie, while the woman is dressed in a blue suit with a pearl necklace. They are both looking serious and are walking side by side. The hallway they are walking in has a red carpet and a statue in the background. The news segment is titled \"Democrats describe 'meltdown' in Syria meeting\". The style of the video is a typical news segment with a focus on the politicians and their reactions to the meeting."
+    ],
+    "video_ids": [
+        "XZLL8ks8dZA_3_256to403"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two politicians, a man and a woman.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two politicians, a man and a woman, standing together. The man is wearing a suit and tie, and the woman is wearing a blue suit with a pearl necklace. The on-screen text identifies the woman as Rep. Nancy Pelosi, Speaker of the House, and the man as a fellow politician, confirming their identities as politicians. Additional people in the background do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two individuals, a man and a woman, standing together. The man is wearing a dark and tie, and the woman is in a blue suit. a necklace necklace. They setting-screen text ' them man as a. S Pelosi, which of the House, which the man as Rep colleague politician, which their roles as politicians. The elements or the background do not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..18c764a1045a9ac301d0069b8ce0a7091382a831
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8efc3fa45a981a097cadfd2274e60e7eee90c7bc949d93f4527961d5c6543dba
+size 179374
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..642354bc43298ff3040fea953ced501b79bb81cb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two men are sitting on a yellow couch, engaged in a conversation. They are both wearing blue shirts with red name tags, suggesting they might be at a formal event or a professional setting. The man on the left is smiling and appears to be enjoying the conversation, while the man on the right is looking down at his hands, possibly deep in thought or contemplating something. The background is blurred, but it seems to be an indoor setting with a black curtain, indicating that they might be in a room or a studio. The overall style of the video is casual and relaxed, capturing a candid moment between the two men."
+    ],
+    "video_ids": [
+        "6__cK3Dg6HI_11_0to181"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men sitting on a yellow couch, engaging in conversation and interaction. Although there is a partial view of a third person on the right, the primary focus and core subject of the video are the two men, fulfilling the 'Object(s): Two men' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two men sitting on a yellow couch, engaging in a. interacting. Both the are a slight view of a third person in the far side the focus focus and the of of the video are the two men, which the 'Object(s)' Two men' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d54f5176a61550cad8423f7d418b188af00130d5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1379-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8831d918ccd4404db7e2d037e05b047967f59c16b5ee48d6fc8a57b86fe4c36d
+size 254543
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..ad161e4b4ba6aeaf2699c82e465df2bbfad766e3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic scene at a skate park. A red BMX bike is the main focus, parked on a ramp. The bike is equipped with a black seat and handlebars, and it's positioned in the center of the frame. The ramp, made of concrete, is located in the background, adding depth to the scene. The skate park itself is filled with various ramps and obstacles, suggesting a space designed for action and stunts. The overall style of the video is energetic and vibrant, capturing the essence of the skate park and the thrill of BMX biking."
+    ],
+    "video_ids": [
+        "t12GSql7DWM_18_0to152"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A red BMX bike? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red BMX bike as the central object, shown in various shots including a full view and close-ups of its wheel. The bike's red frame and BMX characteristics are clearly visible, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a red BMX bike, the main object. which in a angles. close close view and close-ups. the details and The bike is design color and blackX design are clearly visible, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b55a44ff4af4c7d43d805824869aa543a289fd06
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a92077fafdeeedcf46a968588b3b6a67708c2fe455ab8b004cf877866d3ddec3
+size 145106
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..d2ce4af088fe09097804d96b101d03bc552062b2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen in a modern kitchen, leaning casually on a white countertop. He is wearing a red and white striped shirt and a black baseball cap. In his hands, he holds a red smartphone, which he is using. The kitchen is well-equipped with a stainless steel refrigerator and a sink. The man appears to be engaged in an activity on his phone, possibly browsing or texting. The overall atmosphere of the video is relaxed and comfortable, with the man appearing at ease in his surroundings."
+    ],
+    "video_ids": [
+        "Il2zcGI-UR8_11_16to158"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a red and white striped shirt, a black baseball cap, a red smartphone? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a red and white striped shirt and a black baseball cap, holding a red smartphone. These core elements match the description exactly, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing a red and white striped shirt and a black baseball cap. holding a red smartphone. The elements elements match the description provided, and there additional elements are present.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6c18abe2bd7a27eecd46fc5da23f3995155c6b5f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52294217cf1747dbe6f27ec994cf9d882fa32ffac24ae0923cdbff8ac9dc1ed9
+size 130250
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..a769bae804a1e14c9a1273df7befae69cc2e3a13
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a blue plastic tray containing a variety of food items. The tray is divided into three sections, each filled with different types of food. In the first section, there are two sandwiches stacked on top of each other. The sandwiches appear to be made with white bread and filled with a variety of ingredients, including what looks like cheese and meat.  The second section of the tray contains a selection of fresh fruits. There are several slices of apple, which are a bright red color, indicating that they are ripe. The apples are cut into thin slices and arranged neatly in the tray.  The third section of the tray contains a variety of vegetables. There are several slices of tomato, which are a deep red color and appear to be fresh. The tomatoes are cut into thin slices and arranged in a pile.  The style of the video is simple and straightforward, focusing on the food items in the tray. The camera is positioned close to the tray, providing a clear view of the food items. The lighting is bright, highlighting the colors and textures of the food. The video does not contain any text or additional elements, focusing solely on the food items in the tray."
+    ],
+    "video_ids": [
+        "W_DHfg-0uuw_41_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two sandwiches, slices of apple, slices of tomato? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a blue divided plate with two small sandwiches (made with bread, cheese, and tuna), slices of apple, and slices of tomato. These elements match the description exactly, with no conflicting items present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two clear container container with two sandwiches sandwiches onone with white, cheese, and tomato or slices of apple, and slices of tomato. The elements match the description of, fulfilling no additional additional present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c9a1d982bc3869249cf67823793b8e2a354606fd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0662f4901575264f092275416768e2ea9bb63b2f085089ab3f6d7cb218878f2
+size 95958
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..dfdac9dbf9f15d7c188086e5d298952a898cbb82
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows the interior of a car, specifically focusing on the back seat area. The car appears to be a modern SUV or crossover vehicle, with a spacious and well-designed interior. The back seat is folded down, revealing a flat surface that extends from the back of the front seats to the rear of the vehicle. The interior is well-lit, with natural light coming in through the windows, suggesting that the car is parked outdoors. The car's interior features a combination of black and white colors, with the seats and door panels being white, and the floor and other surfaces being black. The car's design includes a variety of storage compartments and cup holders, and the seats are equipped with headrests and seat belts. The overall style of the video is clean and minimalist, with a focus on the car's interior design and functionality."
+    ],
+    "video_ids": [
+        "NCf-uSDEr8Y_26_84to229"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Back seat (folded down), seats, door panels, floor, storage compartments, cup holders, headrests, seat belts? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows the back seats folded down, revealing the cargo area. It clearly displays the seats, door panels, floor, and storage compartments. Seat belts are visible on both sides, and headrests are present on the rear seats. The video does not contradict any of the listed elements, and all are accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows the back seat being down, revealing the floor area. The also displays the seats, door panels, floor, and storage compartments. The belts are not, the sides of and therests are present on the seats seats. The video does not contradict the of the mentioned elements, and the are visible depicted.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0abde0aeb3b692829c2149f8c3cdfe8e7174875b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2b04aba858a3d50493c285ce1959a6359d105972f1503774f77adfa203404b1
+size 209517
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..08a4c614ed5b23d835518bba51f795f274770b32
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up of a boat's dashboard and steering wheel, showcasing the luxurious design and high-quality materials. The dashboard features two round gauges, likely for speed and fuel, with a digital display in the center. The steering wheel is made of polished metal and leather, with a logo on the top. The boat is on the water, as evidenced by the view through the windshield, which shows a serene body of water and a clear sky. The style of the video is a professional product showcase, with a focus on the boat's interior and the craftsmanship of the controls. The lighting is bright and even, highlighting the details of the dashboard and steering wheel. The video is likely intended for promotional purposes, to demonstrate the boat's features and appeal to potential buyers."
+    ],
+    "video_ids": [
+        "zx3lTW4C3BA_20_91to214"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dashboard with two round gauges and a digital display, steering wheel made of polished metal and leather with a logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a boat dashboard with two round gauges and a digital display, along with a steering wheel that is made of polished metal and leather. The logo 'HARRIS' is visible on the dashboard, matching the description. No conflicting elements are present that contradict the specified object conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a dashboard's with two round gauges and a digital display. which with a steering wheel that is made of polished metal and leather, The steering onWornIS' is visible on the steering, which the description. The additional elements are present in would the given objects conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dda1d74188a8e7a41bfaac33dfb16e5a9b6e24b5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f1cf0e3a6b1da691c39b8f739fcb31ae26310a6d72325dd2fc483cd0ea8f4ef2
+size 227103
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..a10a8ae025fc89e063a911c2161eb055ee83cf45
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a silver Hyundai SUV parked in a parking lot. The car is positioned at an angle to the camera, allowing a clear view of its front and side. The vehicle is shiny and appears to be in good condition. The parking lot is made of concrete and has yellow lines marking the parking spaces. In the background, there is a brick building with a window. The style of the video is straightforward and documentary, with no additional effects or filters applied. The focus is on the car, and the background is kept simple to avoid distractions. The lighting in the video is natural, suggesting that it was taken during the day."
+    ],
+    "video_ids": [
+        "JrUHDv4jPUo_5_0to130"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A silver Hyundai SUV? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a silver Hyundai SUV parked in a lot, matching the description. The vehicle's color, model, and brand are accurately represented without any contradictory elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a silver vehicle SUV parked in a parking. which the description provided The vehicle's design, model, and setting are consistent represented, any conflicting elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..723c3f28baad3fd63889046f698ca2134bc2560c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13bcdf25e158ff0b2bf6cca7aa9303120908067e4606388570df748bcb873f05
+size 219712
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..d139a569eff4eacf46e8b672b28516d276c12e84
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men sitting on a bench in a gymnasium. The man on the left is wearing a black leather jacket and a blue shirt, while the man on the right is wearing a gray sweatshirt. They are engaged in a conversation, with the man on the left gesturing towards the man on the right. The gymnasium has a blue and orange color scheme, and there are bleachers in the background. The style of the video is candid and informal, capturing a moment of interaction between the two men."
+    ],
+    "video_ids": [
+        "QA6fJdYXweY_11_0to145"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men sitting on a bench.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men sitting on a bench, which matches the 'Object(s)' condition. The background and their attire are additional elements but do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two men sitting on a bench. which ful the descriptionObject(s)' condition. The men and the attire are consistent elements, do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4400757b142e25d6e9bba85107ab4ce3cd67fa6a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:956a327236c799b62b15cb6014491bf00e6b5b9eb982c1b1b771eb4eff8b8ada
+size 119954
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..ac42bfc097c1ea6bf591f98c9a9ebcff61aaf98f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene underwater scene featuring two fish swimming gracefully through a shallow, clear body of water. The water is teeming with aquatic plants, predominantly pink and green, which sway gently with the current. The fish, one dark purple with lighter markings and another white, move fluidly, their bodies gliding effortlessly past the submerged vegetation. The camera remains steady throughout, providing a continuous view of the tranquil environment. The fish's movements create subtle ripples in the water, adding to the peaceful ambiance. The overall setting evokes a sense of calm and natural beauty, highlighting the delicate ecosystem beneath the surface."
+    ],
+    "video_ids": [
+        "bf6bac38dce49df3afb8df7e7045521287afa5b6c9fe5093c93341318b8726b1"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two fish: one dark purple with lighter markings and another white.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two fish: one that is dark purple with lighter markings and another that is white. Both fish are visible throughout the video, swimming in the clear water among aquatic plants, matching the described condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two fish: one dark is dark purple with lighter markings and another that is white. The fish are present in the video, and among an aquarium blue among the plants. which the description '.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dc729aa10e8fdfb89f549eef8b2be670ae71e945
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-138-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e2fd9412b379fadaa467cb6c9cdf1f4c794455a151041f26817a4ef5f58d37e
+size 236639
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..2cc3bd18a694310ee0214ca17417ed8eeee73f68
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a young man with long hair and glasses, wearing a black t-shirt. He is seated and appears to be engaged in a conversation with someone off-camera. The man's expression is serious, and he is looking to his left, suggesting that he is actively listening or responding to the other person. The setting is a simple, uncluttered room with a white wall in the background, which provides a neutral backdrop that focuses attention on the man. The lighting is soft and even, casting gentle shadows and highlighting the man's features. The overall style of the video is straightforward and unembellished, focusing on the man and his interaction rather than the surroundings."
+    ],
+    "video_ids": [
+        "aXnLWPldLQI_38_0to189"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man with long hair and glasses, wearing a black t-shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man with long hair tied back, wearing glasses and a black t-shirt, which matches the description. The background and other elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man with short hair and back, wearing glasses, a black t-shirt. which align the description provided The presence and other elements in not contradict the description description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bc68a16195b69d4ebc70fee7d4dbb53fbd341508
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:36fd67dcec18e85f8bfa81d53d636c00fd89065ec6a654bae830f98ccf443cce
+size 121058
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..62e66fe46bd182f9e20aff96c908c4d0358c266e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person playing with a toy claw machine. The claw machine is blue and has multiple compartments. The person is holding a small toy in their hand, which appears to be a character from a popular children's show. The toy is being held up to the claw machine, suggesting that the person is about to use the claw to try and win the toy. The background is a bright blue color, which contrasts with the blue of the claw machine. The overall style of the video is playful and fun, with a focus on the interaction between the person and the toy claw machine."
+    ],
+    "video_ids": [
+        "24LdEtCLNfM_16_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue toy claw machine with multiple compartments, a small toy (a character from a popular children's show) held by a person.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a blue toy claw machine with multiple compartments, and a small toy character (a Paw Patrol character) is being held by a person's hand. These elements match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a blue toy claw machine with multiple compartments, which a small toy character froma character Patrol character) being being held by a person. hand. The elements match the description provided,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cc221eda8e16490c1186e2ee15e9aaf687587b0b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bfb1b33e3e204d7a61c626e205ebae8a57f59b1d3c445f89e88aaf91080d1041
+size 163287
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..b93ecf2276f008535596ab598a3b8b86d320b5fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a brown bear is seen playing a black grand piano in a room with a warm, cozy atmosphere. The bear is standing on its hind legs, using its front paws to press the piano keys. The piano is situated on a rug, and there are books and trophies scattered around the room, suggesting that the bear might be a musician or an intellectual. The room has a rustic charm, with wooden walls and a brick fireplace. The bear's fur is a rich brown color, and it appears to be enjoying itself as it plays the piano. The video is likely a 3D animation, given the level of detail and the realistic depiction of the bear and the room."
+    ],
+    "video_ids": [
+        "3oe90vUfPRI_61_29to209"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A brown bear, a black grand piano? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly features a brown bear interacting with a black grand piano, which matches the 'Object(s)' condition. The bear is seen approaching, opening the piano lid, and preparing to play, while the piano remains prominently in the scene. Additional elements like the room decor do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a brown bear playing with a black grand piano. which ful the 'Object(s)' condition. The bear is positioned playing and sitting the piano,, and playing to play, indicating the piano is stationary in the frame. The elements like the fireplace's and not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4152debc4937c490160e549625ac93e0fb5f86b4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:161f68157f8b3963939108b2dcc37c09c722231c407cbfe4ffeeeb188a63fd46
+size 190102
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..a4bdac810d93a1190c0290c9c08d29b092313d13
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are seen in a store. The man is wearing a blue shirt with a white logo on it, and the woman is wearing a red tank top. They are both smiling and appear to be enjoying their time together. The store has shelves filled with various items, and there is a sign that reads \"www.bn.com\". The man and woman are standing in front of a display that has a blue chair and a black bag on it. The woman is leaning on the man's shoulder, and they both seem to be looking at something off-camera. The overall atmosphere of the video is casual and friendly."
+    ],
+    "video_ids": [
+        "6amypbuookU_5_59to256"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a man and a woman as the main subjects, with the man seated and speaking while the woman leans beside him. Although there is another woman in the background, the core description of 'A man and a woman' is clearly fulfilled by the two individuals in the foreground, and no element contradicts this basic requirement.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows features a man and two woman, the main subjects. with the man standing in the, the woman is in him. The there is a woman in the video, the primary focus of 'A man and a woman' is fulfilled fulfilled by the main individuals in the foreground.\"\n making the significant contradicts this.\"\n condition.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..66b1eab36e34c11c0c8babf5ab97420a70997f67
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e54a8c5cb1434ee61d1ad6bdd635adf584ef59794063201e45f5b698defd2a20
+size 195303
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..f588a76ab24e4945b2e993b58b2a32bac3b34363
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a workshop, wearing a blue shirt and glasses, with a beard and mustache. He is holding a wooden object in his hands. The workshop is equipped with various tools and equipment, including a drill press and a vice. The man appears to be in the process of working on the wooden object. The workshop has a patriotic theme, with an American flag hanging on the wall. The man seems to be focused on his work, and the overall atmosphere of the video is one of concentration and craftsmanship."
+    ],
+    "video_ids": [
+        "ZHr8uqUT4bA_0_0to106"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue shirt and glasses, a wooden object, a drill press, a vice? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a blue shirt and glasses, holding a wooden object. In the background, there is a drill press and a vice visible, matching the described objects. The scene is consistent with the 'Object(s)' condition without contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a blue shirt and glasses, holding a wooden object. The the background, there is a drill press and a vice,, which the description elements. The setting appears set with the providedObject(s)' condition provided any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..87736b876deae480f2c64926c9af62c9ce81ca6d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e9337ccb5f2579735eff007ca191e4073119e14c05768701fcbf70230446687c
+size 190240
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..95c546afbf7ed986c7495c7f1806ae728468ff73
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man standing in front of a television screen. He is dressed in a suit and tie, suggesting a formal or professional setting. The television screen displays a logo, indicating that the video may be related to a news broadcast or a television show. The man appears to be speaking or presenting, as he is looking directly at the camera. The overall style of the video is professional and polished, with a focus on the man and the television screen."
+    ],
+    "video_ids": [
+        "Pp2AHvktgeY_2_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, a television screen? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a suit (though no tie is visible) standing next to a television screen displaying the 'sbt CIDADE' logo. The core elements described \u2014 a man in a suit and a television screen \u2014 are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed a suit anddark the tie is visible) in in to a television screen. a wordSTYLET''' logo. The man elements of in a man in a suit and tie television screen \u2014 are present in match depicted in}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1f9beb0ad1b1dc3364dceb812aae79f738f487ec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b9b2c15258c38879591fdc5fc845ceb1ed343f007f87c7b2dd7dd60d894d40b
+size 133944
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..4438736f4f766edaae08c4f40143fe1de0504f1e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a culinary tutorial featuring a pink smoothie. The style is clean and modern, with a focus on the vibrant color of the drink. The presenter, Sandra Matarazzo, is not visible in the video. The production and editing are done by Receitas e Tempos. The video shows a close-up of the pink smoothie in a clear glass, with a white straw and a white spoon. The smoothie is placed on a wooden table, and there are slices of pink grapefruit and green limes in the background. The overall atmosphere of the video is fresh and appetizing, inviting viewers to try the recipe."
+    ],
+    "video_ids": [
+        "GRCK_NcjFmw_21_16to160"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Pink smoothie, clear glass, white straw, white spoon, pink grapefruit slices, green limes? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a pink smoothie in a clear glass, with a white straw inserted. Two white spoons are visible on the wooden surface, and there are pink grapefruit slices (one cut in half) and green limes in the background. All specified elements are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showcasing showing a pink smoothie in a clear glass, a a white straw and. There pink spoonons are also, the left surface, and there are pink grapefruit slices andone whole in half) and green limes in the background. The these objects are present, match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a1df6b8c4762310441377db1fc2cb583c9703903
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f2f0f1dd2d21ff88ed01f1c3d6ac49e525658423b3e681b5b42d786df045e7b6
+size 118855
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..49f8b7ce9966dbb997d220fe217840fa321f6b42
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a playful interaction between three horses in a snowy field. The horses, two white and one brown, are seen frolicking in the snow, their hooves kicking up small clouds of snow. The horses are positioned in a triangular formation, with the white horses on either side and the brown horse in the middle. The snow-covered field provides a stark contrast to the horses' vibrant colors, creating a visually striking scene. The horses' playful behavior suggests a sense of camaraderie and joy, as they engage in their snowy adventure."
+    ],
+    "video_ids": [
+        "EHCpuDHS0FY_13_0to102"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three horses: two white and one brown.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple horse figurines, with two prominently white horses in the foreground and at least one brown horse visible in the background. While there are other horses (including black and white ones), the core description of \\\"three horses: two white and one brown\\\" is largely fulfilled, as the two white horses and one brown horse are clearly visible and central to the composition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two horses figuresines, with two of white and and the foreground and a least one brown horse in in the background. The the are additional elements inor a ones brown ones) the core description of 'three horses: two white and one brown\\\" is largely fulfilled. as the focus white horses and one brown horse are clearly visible and central to the scene.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cdfb4ae6b3b5bc4e6b096aa3f289de78d81b77fe
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1380-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:237327688ab4535d0130f3d344898aecf7ddb835e04299cff559c34a9261d855
+size 370015
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..10c959f8c7883f2202566152e51438cff91498e6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a breathtaking aerial view of a mountainous landscape. The scene is dominated by a series of majestic green mountains, their peaks reaching towards the sky. The mountains are interspersed with serene lakes, their calm waters reflecting the surrounding scenery. A winding road, like a ribbon, traces its way through the landscape, adding a sense of scale and human presence to the otherwise untouched wilderness. The video is a testament to the beauty of nature, showcasing the grandeur of the mountains and the tranquility of the lakes. The aerial perspective provides a unique viewpoint, allowing the viewer to appreciate the vastness of the landscape and the intricate patterns formed by the mountains and lakes."
+    ],
+    "video_ids": [
+        "KPXWLOnFJms_39_61to184"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Majestic green mountains, serene lakes, a winding road? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by prominently featuring majestic green mountains, serene lakes, and a winding road. The mountains are lush and expansive, the lakes are calm and reflective, and the road snakes through the landscape, all of which align with the described elements. No conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition by depicting featuring majestic green mountains, serene lakes, and a winding road. The image are lush and green, the lakes are calm and reflective, and the road me through the landscape, all of which are with the description elements. The additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6041864d2c6d5d963588377c51a0a4e9baae7cb2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ee3383f8f932ac06bdb4485f7edcde0c5e247f2b16a1b96ca425eb1df5f2d2f
+size 237188
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..6ef7bc474c994ba5db3be8d694ba4702a90f38ca
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a muffin tin filled with freshly baked muffins. A knife is inserted into one of the muffins, indicating that they are ready to be served. The muffins are golden brown and appear to be filled with a green filling, possibly spinach or another vegetable. The muffin tin is placed on a cooling rack, which is situated on a wooden table. The table is covered with a white cloth, and there is a blue towel in the background. The style of the video is simple and straightforward, focusing on the food and the process of serving it. The lighting is bright and even, highlighting the texture and color of the muffins. The overall impression is one of a homey, inviting kitchen scene."
+    ],
+    "video_ids": [
+        "ipVccpr3GUM_34_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Muffin tin filled with freshly baked muffins, a knife, muffins (golden brown, green filling), cooling rack, wooden table, white cloth, blue towel? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a muffin tin filled with freshly baked muffins that appear golden brown with green filling, consistent with the description. A knife is being used to lift one of the muffins. The muffin tin is placed on a cooling rack, which rests on a wooden table. A white cloth and a blue towel are visible in the background. All elements described in the condition are present and do not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a muffin tin filled with freshly baked muffins that have golden brown with a filling. a with the description. A knife is used used to cut one of the muffins, The presenceins tin is placed on a cooling rack, which is on a wooden table covered There white cloth and a blue towel are also in the background, The elements in in the condition are present and match not contradict the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cac813e7dda8f012d50205baa9ef4c9c432cfc8b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ac2837efdf8fe29ea7560f20a8f546e2a2c37484727fb9adf1fd3c2b6bc6eeb
+size 199178
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..47aa47cf572bcc18528fc1a232146d85a85c2a89
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures an underwater scene teeming with marine life and vibrant coral formations. The camera glides smoothly over a lush coral reef, showcasing a variety of corals in different shapes and sizes, ranging from branching to rounded forms. The colors of the corals vary from deep greens to lighter hues, creating a rich tapestry beneath the water's surface.  As the camera moves forward, it reveals a school of fish swimming gracefully among the corals. These fish, characterized by their dark bodies with white stripes, glide effortlessly through the water, adding a dynamic element to the serene underwater landscape. The fish appear to be exploring the reef, occasionally darting between the branches and crevices of the coral formations.  The camera continues its journey, capturing the intricate details of the coral structures and the fluid motion of the fish. The interplay of light and shadow enhances the visual appeal, highlighting the textures and contours of both the corals and the fish. The overall atmosphere is one of tranquility and natural"
+    ],
+    "video_ids": [
+        "1e621229ce3128cb4de65fdb597128e0176a92eab82ffed9af13bd16bcf49fbc"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Corals, fish? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows corals and fish as requested. The underwater environment is filled with various types of coral formations, and a fish is clearly visible swimming among them. The presence of other marine life and coral structures does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful corals and fish, described. The image scene is depicted with vibrant cor of cor,, and there school is visible visible swimming among them. The presence of additional elements life or the types does not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e0f2395c278b43dd68bb6ea7a9524794305f7090
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0da8c56ca6d70a5c97bb3852db6bc3aa2a6b79142b7b32321a4571d712eb870c
+size 286887
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..fc5266de7d1ba48c416233c54312b6602b134089
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a butterfly perched on a green leaf. The butterfly, with its vibrant orange and black wings, is the main focus of the video. It is seen in three different positions: facing the camera, facing away from the camera, and facing the camera again. The background is blurred, emphasizing the butterfly and the leaf it is perched on. The video is shot in a way that showcases the butterfly's beauty and the intricate details of its wings. The overall style of the video is a close-up nature shot, highlighting the butterfly's colors and movements."
+    ],
+    "video_ids": [
+        "2jwvdVUQGRQ_23_0to138"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Butterfly? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a butterfly with distinct orange, black, and white wing patterns resting on a green leaf, which matches the 'Object(s): Butterfly' condition. The butterfly's features and behavior are consistent with the description, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video consists depicts two butterfly in vibrant orange and black, and white patterns patterns, on a green leaf. which align the descriptionBut(s)' Butterfly' condition. The presence is detailed, the are consistent with the description, and there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..97089dcb9476b890d471e59a3e4c2245079c5dc3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7d1bf8a9b8beae21587b1466864b016e86a820ce89a41abb8056704f0ede15b8
+size 203576
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..7777768451015331a5c0a22d6438f59546330d6e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a candid snapshot of a man and a woman posing together at a pool party. The man, wearing a blue baseball cap, is smiling and looking directly at the camera, while the woman, with her long eyelashes and full lips, is also smiling and looking slightly to the side. The background is filled with people enjoying the party, and the pool is visible in the distance. The overall style of the video is casual and fun, capturing a moment of joy and relaxation."
+    ],
+    "video_ids": [
+        "1Jl_rBysMKM_10_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: - A man? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a blue baseball cap with 'LA' on it and a black polo shirt. He is positioned next to a woman, and the background appears to be a poolside setting with people and umbrellas. The man is the central focus of the video, and his presence fulfills the 'Object(s): A man' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man with a blue cap cap and aB' written it, a beard shirt shirt. The is positioned next to a woman, and both setting suggests to be a socialside setting with other in greenrellas, The presence is the primary figure of the video, fulfilling his presence fulfills the 'Object(s)' A man' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..eccedfb9505e9ec9d68ec3bc77d948bd1adfc178
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e9e532c6fdbce8090e151aeed6fbee30d4f3bb74a00431f9dc48330ffb7cd18
+size 199774
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..efb74d347a8912a4499e11caf175f581056ed381
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman is seen lying in bed, pulling the covers over her head. She is wearing a black bracelet and has a ring on her finger. The bed is adorned with blue and white patterned pillows. A white letter A is hanging on the wall behind her. To her right, there is a lamp with a white shade and a silver base. A vase with white flowers is also present on the bedside table. The scene suggests a cozy and comfortable setting."
+    ],
+    "video_ids": [
+        "GxY6Y0wHcfc_3_0to134"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Woman, black bracelet, ring, white letter A, lamp with white shade and silver base, vase with white flowers, bedside table? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows a woman in bed, wearing a black bracelet and a ring, with a white letter 'A' on the wall, a lamp with a white shade and silver base, a vase with white flowers, and a bedside table. All described objects are clearly visible and consistent with the provided description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a woman lying a, wearing a black bracelet and a ring. with a white letter 'A' on the wall behind a lamp with a white shade and a base, and vase with white flowers, and a bedside table. The these elements are present visible and match with the video description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..664559f2435d82ca2736d7b5f0dfe12cd7219318
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0dad62ee670d859417e35f422d0177ab0d5de9698ab4b9ba851d1d77e22d3f34
+size 268687
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..43bc5811cbc2f568f8a89e9d03ea3eb730f01175
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures an underwater scene featuring a group of sockeye salmon swimming through a rocky riverbed. The water is clear enough to see the pebbles and stones at the bottom, which vary in size and color, ranging from light gray to dark brown. The salmon are vibrant red with hints of purple on their bodies, and they move gracefully through the water, their fins and tails creating gentle ripples. As the video progresses, more salmon come into view, swimming in various directions, some close together while others maintain a bit of distance. The lighting is natural, likely from sunlight filtering through the water above, casting a soft glow on the fish and the riverbed. The overall atmosphere is serene and natural, highlighting the beauty and movement of these fish in their aquatic habitat."
+    ],
+    "video_ids": [
+        "5ec2a221bb66569d9490a5b727ec94afbc5515bd4b2ad651111b4343331d9034"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Sockeye salmon? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows sockeye salmon, identifiable by their distinctive red bodies and greenish heads, swimming in a riverbed with rocks. Their movement and appearance match the description of sockeye salmon, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts fisheye salmon, which by their vibrant red color and whiteish-blue, swimming in an clearbed. pe. The appearance and appearance align the description of sockeye salmon, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cc2ad39762a15d724639b6795e6e16575ab1f77a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:862a7552aa601f7dc7730e87054f6274c9df007422ef291ea1e9cefc22d6d429
+size 272720
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..2a9b8168ee05f6a05b8f3cec2d8613a306589071
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a waterfall cascading into a pond. The waterfall, with its mossy rocks and lush greenery, is the focal point of the image. The water, a deep shade of green, is dotted with small fish, adding a touch of life to the scene. The waterfall's flow is steady, creating a soothing sound that fills the air. The surrounding area is teeming with plants and trees, their leaves a vibrant mix of green and brown. The overall style of the video is naturalistic, capturing the beauty of nature in its raw form. The camera angle is low, looking up at the waterfall, which adds a sense of grandeur to the scene. The video is a testament to the tranquility and beauty of nature."
+    ],
+    "video_ids": [
+        "-fU69dMu8dU_49_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Waterfall, mossy rocks, lush greenery, small fish, plants, trees? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a waterfall cascading over mossy rocks, surrounded by lush greenery including various plants and trees. Small fish, likely koi, are visible swimming in the pool at the base of the waterfall, matching the described elements. The scene is consistent with the requested objects without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a waterfall cascading over mossy rocks, surrounded by lush greenery. plants plants and trees. The fish can possibly koi, are swimming swimming in the water below the base of the waterfall. which the description elements. The presence is vibrant with the ' ' and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a4e3a4d7807d6eab05d06b615aff4ea8471f0741
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1381-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89a0692ef6060ae73d2254111ec4ff730a268cbc2475edc6416aca7011b81635
+size 193786
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..5176e39e6c54e2c5a15d991a90a65cb711baa0c1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with blonde hair and glasses, who appears to be in a relaxed state. She is wearing a black top and is seated in a room with a blue wall. The room has a modern design, with a white countertop and a painting hanging on the wall. The woman is the main focus of the video, and her facial expressions suggest that she is either yawning or closing her eyes. The overall style of the video is casual and intimate, capturing a candid moment in the woman's life."
+    ],
+    "video_ids": [
+        "AZRrnkh77bU_2_0to112"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with blonde hair and glasses, wearing a black top, and seated in the room.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair and glasses, wearing a black top, and seated in a room. All core elements of the description are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair and glasses, wearing a black top, seated seated in a room. The the elements of the description are present, accurately depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8d61544804d9870d59fc76d7bdcf23f915042343
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:44c2cc4aa2d5ecfef79929713a27f74ac970256a3038de6d5fb6528a7f87ef28
+size 114409
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..cfee5ade922cc8451f18b62b1693fd8d3244a142
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features two men standing in a motorcycle gear store. The man on the left is wearing a purple t-shirt with a graphic design on it, while the man on the right is wearing a black and white motorcycle jacket. They are standing in front of a display of motorcycle gear, including helmets and jackets. The store has a modern and stylish design, with a red and black color scheme. The men appear to be engaged in a conversation, possibly discussing the motorcycle gear on display. The overall style of the video is casual and friendly, with a focus on the motorcycle gear and the store itself."
+    ],
+    "video_ids": [
+        "oTKHnJYU-H8_1_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, a display of motorcycle gear (helmets and jackets), a purple t-shirt with a graphic design, a black and white motorcycle jacket.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men standing in front of a display of motorcycle gear, including helmets and jackets. One man is wearing a purple t-shirt with a graphic design, and the other is wearing a black and white motorcycle jacket. All core elements described are present and accurately represented in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts two men standing in a of a display of motorcycle gear, including helmets and jackets. One of is wearing a purple t-shirt with a graphic design, and the other is wearing a black and white motorcycle jacket. The elements elements of in present and accurately depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b9fdc3456c342626916bd2fd069edc74ed9f7bca
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1fd00b9c66281dfc862c58b129cb7236cbfd5c140e53505d684397018c6dc4c
+size 193546
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..f086596922cdb45a787c696ff55d21d3907db7d2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a black Method Race Wheels wheel in three different stages of rotation. The wheel is designed with a unique spoke pattern and a black center cap that features the brand name. The style of the video is a simple product showcase, focusing on the wheel's design and craftsmanship. The wheel is the only object in the video, and it is presented against a plain white background to avoid any distractions. The video is likely intended for promotional or sales purposes, highlighting the wheel's features and design."
+    ],
+    "video_ids": [
+        "d5ld4F-Giqs_11_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Black Method Race Wheels wheel? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a single black Method Race Wheels wheel, which matches the description. The wheel is clearly visible, and no conflicting elements are present that contradict the object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a close black wheel Race Wheels wheel spinning which matches the description provided The wheel is stationary visible, and the additional elements are present. would the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dacf9ac1ac0edeaf574605a1af445394068b9296
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:47547b5b6c4019462ec3646145edc032f148e8ac7d60a475b763d5711a8d1863
+size 295050
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..d6fde6eaaafc25c17fbac12bdd07e47e869ef750
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a single bird wading through shallow water. The bird, likely a shorebird given its long legs and slender beak, is seen moving slowly across the surface of the water. The water is dark and reflective, with small ripples and patches of light creating a shimmering effect. The bird's movements are graceful and deliberate, as it steps carefully over the water, occasionally dipping its head to search for food or to interact with the water. The background is mostly obscured by the water's surface, but hints of a darker, possibly wooded area can be seen in the distance. The overall atmosphere is calm and peaceful, with the bird's actions being the primary focus of the video. There are no other significant objects or characters present in the scene, and the camera remains stationary throughout, allowing the viewer to fully absorb the tranquil beauty of the moment."
+    ],
+    "video_ids": [
+        "36f3ba2b8cee8a3e2122de064282e13bb359b4e6b054cc49d9243819c9810748"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A single bird, likely a shorebird with long legs and a slender beak.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a single bird with long legs and a slender beak, consistent with a shorebird, wading through shallow water. The bird's movements and appearance match the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a single bird, long legs and a slender beak, which with the shorebird. standingading in a water. The bird's white and the align the description provided.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d40ed5ecdc432c840b8f5475f1fbaa5be433b5ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:580e25fdb0c12f9ca22a7bdc388bbb5dc0493a7c38e6496e8e4e7ffc58df0fe0
+size 267358
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..a16df37c516c3d7264d0e0b74e57013e48ef67ff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a small white sailboat gliding across a clear, turquoise lagoon. The boat, with its white hull and mast, is the main focus of the video, appearing in all three frames. The first frame shows the boat sailing towards the camera, its sail catching the light as it moves. In the second frame, the boat is seen from a distance, its small size contrasting with the vastness of the lagoon. The third frame shows the boat sailing away from the camera, its sail billowing in the wind as it moves further into the distance. The lagoon itself is a stunning shade of turquoise, its calm waters reflecting the clear blue sky above. In the distance, a small island can be seen, adding to the sense of tranquility and isolation. The video is shot from a high angle, providing a bird's eye view of the boat and the lagoon, and emphasizing the vastness of the landscape. The overall style of the video is peaceful and serene, capturing the beauty of nature and the simple joy of sailing."
+    ],
+    "video_ids": [
+        "Kt3x9xrg_EY_0_0to168"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small white sailboat? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a small white sailboat as the central subject, floating on turquoise waters with a lush, mountainous island in the background. The boat is clearly visible and matches the description. Additional elements like another boat in the distance and underwater features do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a small white sailboat as the central object. which on clear waters. a clear green greenous backdrop in the background. The sail remains consistent visible in matches the description of The elements like the sail in the distance and the vegetation do not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6924fb39c337d0e6ab1feabbdc99a0c78a36d461
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11aef4d292f9386320383a2774997a0fc22103eb6c8e41e980f1b911fab81179
+size 263597
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..14ec283a2870a6ea72b6dac1964667c5515be3d6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a woman in a purple shirt standing in a fabric store. She is holding a piece of fabric and appears to be speaking or explaining something. The store is filled with shelves of colorful fabrics and patterns. The woman is the main subject of the video, and her actions suggest that she might be a customer or a store employee. The style of the video is casual and informative, possibly aimed at providing viewers with information about the fabrics or the store itself."
+    ],
+    "video_ids": [
+        "PaZqrdoi6zI_33_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a purple shirt holding a piece of fabric.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a purple shirt and holding a piece of fabric. The core description is accurately represented, with no contradictions. Additional elements such as shelves of fabric and a table with sewing tools are present but do not conflict with the main subject.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman in a purple shirt, holding a piece of fabric. The fabric description is met represented in with the additional or The elements, as the with colorful in a colorful are more materials are present but do not conflict with the main description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f3546f5e6f158bde38f2837380d5b348f722b914
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4ee1d960ddeb9a5769c220655f6181b1f3e56a914d9a6d8be4059cbce9d4df77
+size 234293
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..819ef8cf446e65882a417decdd2362f3f07386a3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment from a basketball game. The main focus is on two players, one wearing a maroon jersey and the other in a white jersey. The player in the maroon jersey is looking off to the side, possibly strategizing or observing the game. The player in the white jersey is looking down, possibly at the court or his shoes. Both players are on the court, surrounded by the typical elements of a basketball game, such as the hoop and the court markings. The style of the video is a standard sports video, capturing the intensity and focus of the players during the game."
+    ],
+    "video_ids": [
+        "Zk4cxa5NTl8_1_20to155"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two players, one in a maroon jersey and one in a white jersey.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two basketball players. The player in the foreground is wearing a maroon jersey with \\\"CLEVELAND\\\" visible, and the player in the background is wearing a white jersey. This matches the description of two players, one in a maroon jersey and one in a white jersey.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two players players, One player on the foreground is wearing a maroon jersey, theHVELAND\\\" written on while the player in the background is wearing a white jersey with The matches the description of two players, one in a maroon jersey and one in a white jersey.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c42ecea127f8cc9fd42859fc4ad9bd9dd6a301c6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86eeafdf4fa0e04ed630ce8a15a487eeeacc85b75d49876455ee5fad3ace6f75
+size 141136
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..a75bb376d1a85525b411471657bd7ffa18cb8b72
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a 3D animated depiction of a tropical jungle scene. The sun is setting, casting a warm glow through the dense foliage. The viewer's perspective is from the ground, looking up towards the sky, which is visible through the gaps in the trees. The jungle is lush with various types of trees and plants, and a small stream meanders through the scene. The overall style of the video is realistic with a focus on natural lighting and textures."
+    ],
+    "video_ids": [
+        "tvgV-oLFCfI_8_0to171"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Trees, plants, a small stream? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by prominently featuring trees and plants, along with a small stream flowing through the scene. The lush vegetation and water body are clearly visible throughout the video, matching the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting featuring trees and plants, which with a small stream. through the scene. The lush green and the body are clearly visible, the frames, which the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..83b7930e794e0a02c5c53008b4038f81cfecbdb9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1382-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f19048a159c0c93605d79f7d3fefe62daa8e4072d6c6a5c28e29f36fbcba254b
+size 268016
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..df50e8fe7c7c9ee17b972a2cae1e1d47d7e7f032
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a news anchor named Keith Cate. The style of the video is a professional news broadcast with a focus on a specific topic. The news anchor is dressed in a suit and tie, and he is standing in front of a news desk with a backdrop that includes a large screen displaying a news headline. The headline reads \"CEMINOLE HEIGHTS MURDERS DEATH PENALTY DECISION\". The video likely includes a discussion or report on a recent decision related to the Ceminoole Heights murders. The overall tone of the video is serious and informative, typical of a news broadcast."
+    ],
+    "video_ids": [
+        "lMYw2ed7A_0_0_0to158"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: News anchor named Keith Cate, news desk, large screen backdrop with a headline.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a news anchor, identified as Keith Cate, sitting at a news desk. Behind him is a large screen displaying a headline about the 'SEMINOLE HEIGHTS MURDERS DEATH PENALTY DECISION'. All core elements of the 'Object(s)' condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a person anchor in presumably as Keith Cate, standing in a news desk. The the, a large screen backdrop a colorful and ' 'LNIO HS'URDERERS'BSALTY CASEISION.' The elements elements of the descriptionObject(s)' condition are present: match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ef77b22b5b076fcae1eafa7b4f52edea649dce5f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b21a601541c996b6cb63d8bf33186d64bb775700a7521828216a7e7e39d5a25
+size 229709
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..ee2bb309c25d052f9caf64d0a62151d8dedee978
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment on a basketball court. The main subject is a basketball player, dressed in a white jersey with blue and yellow accents, and the number 23 prominently displayed. The player is in motion, running across the court with a focused expression. The player's tongue is out, indicating intense concentration or perhaps a moment of triumph. The background is filled with the blurred figures of other players, suggesting a competitive game in progress. The style of the video is a fast-paced, action-packed sports montage, capturing the energy and excitement of the sport."
+    ],
+    "video_ids": [
+        "sKhV_hkI_4A_4_0to170"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A basketball player in a white jersey with blue and yellow accents, numbered 23.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a basketball player wearing a white jersey with blue and yellow accents, and the number 23 is clearly visible on the jersey. The player's attire and number match the description provided, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a basketball player wearing a white jersey with blue and yellow accents, and the number 23 is clearly visible on the jersey. The player is uniform and the match the description provided, indicating the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..95db1e8ad66b3a3635c69c157d6da951c6a3d67b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2527233f7ec42a4cf394619e1953db437e357a7146e71ddaf583541d35ef8848
+size 244629
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..95f0773854b9ec254af5840f86ef8e337922edde
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a hockey game. The main focus is a hockey goalie, dressed in a white and gold uniform with the number 29, standing in front of a red and white net. The goalie is in a crouched position, ready to make a save, with his gloved hands outstretched. The background is filled with the blurred figures of other players, indicating the fast-paced action of the game. The style of the video is a real-time action shot, capturing the intensity and anticipation of the moment."
+    ],
+    "video_ids": [
+        "wE6EkuJ5rt0_3_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Hockey goalie in a white and gold uniform with the number 29.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a hockey goalie wearing a white and gold uniform with the number 29 clearly visible on the jersey. The goalie's gear, including pads and mask, matches the description, and there are no conflicting elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a hockey goalie in a white and gold uniform with the number 29. visible on the back. The goalie is stance, including the and gloves, is the typical. and the are no elements elements in contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e83dee4dd17dc80939a4fe7e85753ca89e77da6e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3cbe5fa0733eedaa97b87868dec44e85d8caa9bf44a6e8f373247a54a4810dc8
+size 362621
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..69d6d140a146b7e0c447c30e9e58fe0362318274
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a black leather jacket and jeans standing next to a red sports car in a showroom. The car is a convertible with a sleek design and a shiny finish. The showroom has a clean and modern look with white tiled floors and a large window in the background. The man appears to be examining the car, possibly considering a purchase. The lighting in the showroom is bright and even, highlighting the car's features and the man's interest in it. The overall style of the video is realistic and straightforward, capturing a common scene in a car dealership."
+    ],
+    "video_ids": [
+        "ZuK4PEj6hQk_1_48to184"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a black leather jacket and jeans, a red sports car (convertible with a sleek design and shiny finish).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black jacket and jeans standing next to a red convertible sports car with a sleek design and shiny finish, matching the described objects. The background elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black leather and jeans, next to a red sports sports car with a sleek design and a finish. which the description elements. The car appears, not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..88f5aafe603c794c1eaecfffe03e43cbd54c28f5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d2f38fe49e856ba6ab5f0e1debba98e6e38d74611eefd2e955f5198a31b19759
+size 152114
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..3f1449d8c9c7901390e5a8679f2c5e068e3def04
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a drone shot of a large, green, rice paddy field. The field is divided into sections, and the rice is in various stages of growth. The field is surrounded by a dirt road and a fence. In the background, there are mountains and trees. The sky is clear and blue. The video is in color and has a wide-angle lens. The style of the video is realistic and it captures the beauty of the rice paddy field and the surrounding landscape."
+    ],
+    "video_ids": [
+        "ltYwYvsr1RM_7_0to185"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Rice plants in various stages of growth? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows expansive green rice fields divided into neat rectangular plots, which are indicative of rice plants in growth stages. The varying shades of green across different sections suggest different growth phases, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a fields fields fields, into terr, sections, which suggests characteristic of rice cultivation in various stages. The varying shades of green suggest the sections of different growth stages, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bbf41e2cf851be235652fb06bbe2e60e54ebb7ac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0af41eb62ffe2ef2538cbc631ba968315c5f73ce8e560d1defad07f36bed030d
+size 134338
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..22863b7a7a1286bc0a5bd45f42db71300e49c3ee
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, an elderly woman and a young man are seated at a wooden table, enjoying a meal together. The woman, dressed in a vibrant red jacket, is seen using a spoon to eat from her bowl. The man, wearing a blue shirt, is also eating from his bowl. The table is set with two bowls, each containing a meal. The background features a brick wall, adding a rustic charm to the scene. The video captures a heartwarming moment of intergenerational bonding over a shared meal."
+    ],
+    "video_ids": [
+        "03ucHF-TN5Y_38_22to194"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: An elderly woman and a young man.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows an elderly woman with gray hair and glasses, wearing a red jacket, and a young man with brown hair, wearing a blue sweater and scarf. Both are seated at a wooden table eating food from bowls, which matches the description of 'An elderly woman and a young man.'\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts an elderly woman and gray hair and a, wearing a red jacket, and a young man with short hair, wearing a blue shirt. a. Both are seated at a table table, from from white, which align the description of anAn elderly woman and a young man.' The}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..86ce5a779ce1d2bf343cd76d5436fc309ba1579f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2149bae9de23ca03f6c25562b1f72ba322318a36037fef86f8d94305a73d6c15
+size 123330
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..7b40d42d66fdd793b84a2e130bf4d8a8807c31b6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a time-lapse of a car's interior, showcasing the dashboard and steering wheel. The car's interior is luxurious, with a wooden dashboard and a large touch screen display. The steering wheel is black and features a logo in the center. The car is driving through a forested area, with trees visible through the windshield. The time-lapse effect creates a sense of motion and progression, as the car moves through the landscape. The overall style of the video is sleek and modern, with a focus on the car's interior and the natural scenery outside."
+    ],
+    "video_ids": [
+        "Bu0UhBrzd8k_13_0to119"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A car with a luxurious interior featuring a wooden dashboard, a large touch screen display, and a black steering wheel with a logo in the center.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows the interior of a luxury car with a wooden dashboard (orange-brown wood trim), a large central touch screen display showing a map, and a black steering wheel featuring a central logo (Rolls-Royce emblem). These elements align with the described 'Object(s)' condition, even though the video does not show a person or any other additional elements that contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a interior of a car car with a wooden dashboard,though-brown color),), a large touch touch screen display, a map, and a black steering wheel with a logo logo.likelys-Royce).). The elements align with the description 'lux(s)' condition, making though the wood does not provide the wooden driving the additional additional elements.\"\n might the core.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0ce00975d4adef4e5aab2145f15e7308421bb89d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3f25b4ac34e6619a5368bd36369661e5af6ecae2982432f916f02712e2733aad
+size 274168
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..ee8bb5d78ed2174c85f5c971d92855bd5caa4b98
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a festive holiday scene featuring a delicious-looking pie as the main subject. The pie is placed on a table adorned with holiday decorations, including pine cones, red berries, and a red and white striped cloth. The pie itself has a golden-brown crust and a rich, dark filling that suggests it could be a pecan or chocolate pie. The table setting also includes a cup of coffee, adding to the cozy and inviting atmosphere of the scene. The overall style of the video is warm and inviting, with a focus on the holiday theme and the mouth-watering pie."
+    ],
+    "video_ids": [
+        "H3pxLfXLCRU_29_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A pie with a golden-brown crust and a rich, dark filling (likely pecan or chocolate), a cup of coffee? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a pie with a golden-brown crust and a rich, dark filling, consistent with a pecan or chocolate pie. A cup of coffee is also clearly visible in the scene, fulfilling the object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a pie with a golden-brown crust and a rich, dark filling, which with a pecan or chocolate pie. Additionally cup of coffee is also visible visible in the background, fulfilling the ' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d722a0e489804d10d2a983cd976feb35e6ee6dbb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1383-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:39d0560b4c8522915d96a8a8c63948b58cf6fbf25fd3fa8ae29e5d3350ca4d68
+size 103238
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..febb1550506c4a3b01ab64efd79c8f4847a47d26
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a young girl with glasses, wearing a white shirt with pink and yellow patterns. She is indoors, surrounded by toys and other objects. The girl is looking directly at the camera, and her expression is neutral. The room appears to be a playroom, with various toys scattered around. The lighting in the room is bright, and the overall atmosphere is casual and relaxed. The style of the video is candid and informal, capturing a moment of the girl's daily life."
+    ],
+    "video_ids": [
+        "2d6K6BMiaUU_8_0to132"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Young girl with glasses, white shirt with pink and yellow patterns, various toys and objects.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a young girl wearing glasses and a white shirt with pink and yellow patterns, as described. The background includes various toys and objects, consistent with the condition. The girl's appearance and the environment match the description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young girl wearing glasses and a white shirt with pink and yellow patterns, which described. The background includes various toys and objects, which with the given. The presence's appearance and the setting match the given provided any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2a27db39d7993e47d1b57a06f7a70caaab600e96
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b5ac8b07fed91b95d2e0868abe86144e2cb8ca01c2e52b51c5d4e7c208976ce
+size 134099
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..c6abf0f927cc7032b704c633a436834346d672ee
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a plate of food being served. The plate is white with blue designs around the edge. On the plate, there are three main components: a serving of shredded vegetables, a serving of crispy chips, and a serving of sliced meat. The vegetables appear to be a mix of greens and other vegetables, possibly including cabbage and carrots. The chips are golden brown and appear to be made from a thin, crispy vegetable, possibly potato or sweet potato. The meat is sliced into thin strips and has a dark brown color, suggesting it may be cooked or grilled. The food is presented in a way that suggests it is ready to be eaten, and the close-up view emphasizes the textures and colors of the different components. The style of the video is straightforward and focuses on the food, with no additional context or background provided."
+    ],
+    "video_ids": [
+        "_Qjp76gwnGM_59_0to107"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Three main components: shredded vegetables, crispy chips, and sliced meat.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows three main components: shredded vegetables (bean sprouts with green onions), crispy chips (golden, thin, fried chips), and sliced meat (grilled, seared meat cut into rectangular pieces). These elements are distinctly visible and occupy the primary focus of the plate, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows three main components: shredded vegetables,which sprouts and some and and crispy chips (yellow fried fried, and noodles), and sliced meat (darkilled or darkared meat with into slices pieces). The elements are distinct visible and match the plate space of the plate, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..59e0481c064ca074e0e2b8b081df0af19e10b999
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7d970834cc893f6b26cf00c4da93e3a34b25535e693331c35b1cb315afb66472
+size 82503
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..e8fc95579e04b35f162139e1900399628e009352
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a soccer match. The main focus is a soccer player, dressed in a vibrant yellow and red striped jersey, who is in mid-air, celebrating a goal. His arms are raised high, and his mouth is open in a triumphant yell. The player's position and expression suggest a moment of victory and joy.  The background of the video reveals a bustling soccer stadium filled with spectators. The crowd, though blurred, adds a sense of scale and excitement to the scene. The stadium's architecture and the spectators' attire suggest a professional match.  The style of the video is a combination of action and sports photography. The focus on the player in mid-air, along with the blurred background, creates a sense of motion and energy. The colors are vivid, with the yellow and red of the player's jersey standing out against the more muted colors of the stadium. The overall composition of the video suggests a moment of high drama and excitement in a soccer match."
+    ],
+    "video_ids": [
+        "aaLDDeiLgtc_0_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A soccer player in a vibrant yellow and red striped jersey? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a vibrant yellow jersey with red and blue stripes, which matches the description. The player is clearly visible and the jersey's colors and design are consistent with the specified condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a soccer player wearing a vibrant yellow and with red stripes blue stripes, which matches the description of The player is also in on the jersey's colors and design are prominent with the description description.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dd276a55bced247bb5de3dc4e5b4f71bbc0608e0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:327010ee726bc681e53fbdfb18e3f9f983655e229a34d3a8236cdc5bf28315ff
+size 202654
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..daa6d0aff2c4b73e210aee0235a5786309c185c7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a soccer match. The main focus is on a player in a blue and red jersey, who is skillfully dribbling the ball down the field. He is in the center of the frame, with his body leaning forward in anticipation. His eyes are focused on the ball, which is slightly ahead of him, indicating his intent to control it.  In the background, two players in yellow jerseys are in pursuit. They are running towards the ball, their bodies leaning forward in an attempt to intercept it. Their positioning suggests they are working together to defend their goal.  The field is a vibrant green, contrasting with the players' colorful jerseys. The perspective of the video is from the side of the field, providing a clear view of the action. The focus is on the players and the ball, with the background slightly blurred, emphasizing the intensity of the moment.  Overall, the video captures the excitement and skill of soccer, with the players' movements and the ball's trajectory creating a sense of anticipation and excitement."
+    ],
+    "video_ids": [
+        "67cH6swQElc_32_107to252"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a blue and red jersey, a player in a yellow jersey (pursuing), another player in a yellow jersey (pursuing), and a soccer ball.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a blue and red jersey (Barcelona kit) dribbling the soccer ball while being pursued by two players in yellow jerseys (Villarreal kit). All core elements described are present, and there are no conflicting elements that contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts a player in a blue and red jersey dribthecelona)) dribbling a ball ball, being pursued by two players in yellow jerseys.likelyillarreal kit). The the elements of in present: including the are no additional additional. contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dc25707080d9439d06ccd610256ce222b985dea6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e866a59dc2ddcc81ebef0cd9a7c41b0ae4cc766cc092358bba4ae31438be40fc
+size 254268
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..5ac6ca49d979d16aed1537fec4936a77953a6430
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman sitting on a set with a blue background. She is wearing a maroon dress and has her hair styled in a bun. The woman is gesturing with her hands as she speaks, and she appears to be engaged in a conversation. In the background, there is a plant and a cup on a table. The video has a purple banner at the bottom with the text \"UP NEXT More Girl Chat Coming Your Way!\" The style of the video suggests that it is a segment from a talk show or a similar television program."
+    ],
+    "video_ids": [
+        "7TAzMGMz7FU_18_0to167"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Woman, plant, cup, table, purple banner? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a woman, a plant in the background, a cup on the table, the table itself, and a purple banner with text. All specified elements are present and identifiable in the video frame.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a woman, a plant, a background, a cup on a table, and table itself, and a purple banner with text. The these elements are present and do in the video.\"\n.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f4682650eee45bdd10b810a1ab40586bf6a4a13f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e7f22ebfa6f417c7a838c458541c28edcea68c9d8c186d7be241111260deb226
+size 165106
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..440c7ddeab85d958b65b68ba5ca9ad2172d2db86
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases the interior of a red Lexus sports car, taken from the perspective of the driver's seat. The car's interior is sleek and modern, with a black steering wheel and red leather seats. The dashboard is adorned with a variety of buttons and controls, and the car is equipped with a touch screen display. The car is parked in a garage, and the camera pans around the interior, providing a comprehensive view of the vehicle's features. The video is likely intended for promotional purposes, highlighting the car's luxurious and high-tech design."
+    ],
+    "video_ids": [
+        "K-nv1IHPihw_14_0to125"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red Lexus sports car, black steering wheel, red leather seats, dashboard with buttons and controls, touch screen display? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows the interior of a red Lexus sports car with a black steering wheel, red leather seats, and a dashboard with buttons and controls. Although a touch screen display is not explicitly visible, the central console area appears to have modern controls and interfaces typical of a luxury car, which aligns with the description. The presence of additional elements like the 'Kelley Blue Book' logo does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a interior of a Lexus Lexus sports car with a black steering wheel, red leather seats, a a dashboard with buttons and controls. The the touch screen display is not visible mentioned, the presence console area suggests to have a controls that a, of a car vehicle, which coulds with the description. The presence of additional elements like the carLx' Book' logo in not contradict the core description and}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9856dd004bcc6bb9d78f7d96e557c196e8121287
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:74abacbf905c94d9c9a4e12a1dccb07d39439aa613eadf63607b1333314de6fd
+size 143962
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..1799cb541845903d90ae0751b53c831257145588
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a family's journey through an airport. In the first frame, a man with a concerned expression is seen, possibly worried about their flight. In the second frame, a woman and a young girl are seen, the woman looking at the man with a reassuring smile, while the girl seems to be enjoying the moment. In the third frame, the man is seen again, this time with a relieved expression, suggesting that their flight has been confirmed or they have reached their destination. The video is a snapshot of a family's travel experience, capturing the emotions and interactions of the family members."
+    ],
+    "video_ids": [
+        "XOPVanRTBmc_60_0to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: ['a man', 'a woman', 'a young girl']? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man, a woman, and a young girl sitting together in what appears to be an airport waiting area. All three individuals are prominently featured and match the specified objects in the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video contains depicts a man, a woman, and a young girl. together in what appears to be a airport setting area. The three individuals are present featured, are the description '. the description.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..94a2748f5874e939570f600aecc2bcec7af491cc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb5b37644481b215ff650cbd8bb3c38d10ed496bcb4736c4d60ef93da6d0a273
+size 99425
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..c2618e09aaca3b573556a6080bcadcbe97d938c2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of celebration and camaraderie among a group of soccer players. The players, dressed in vibrant blue and red jerseys, are seen high-fiving each other in a display of team spirit and victory. The scene is set on a soccer field, with the players positioned in the foreground, their faces lit up with joy and pride. The background reveals a crowd of spectators, their faces a blur of anticipation and excitement. The video is a dynamic snapshot of a triumphant moment in the sport, capturing the essence of teamwork and the thrill of victory."
+    ],
+    "video_ids": [
+        "lRj8YxzBeYc_41_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Soccer players, spectators? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows soccer players in team uniforms celebrating on a field, with spectators visible in the background. These elements align with the 'Object(s)' condition of soccer players and spectators, and no conflicting elements are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts soccer players in blue uniforms engaging on a soccer, which one visible in the background. The elements align with the 'Object(s)' condition of the players and spectators, making there additional elements are present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cd1d87f3f98d3c67a783b4175797d8c280ee2359
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1384-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4eadc6065fb10f1de4f7b4af1f2cafa57a5649edf5a59a5dfe1cf4d101366676
+size 250768
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..f85cb430677aba2ce729c00fc318e2768c424932
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the stunning view of the Burj Al Arab, a luxurious hotel located in Dubai, United Arab Emirates. The hotel, known for its distinctive sail-shaped architecture, stands tall against the backdrop of the clear blue sky. The video is taken from a distance, allowing the viewer to appreciate the grandeur of the hotel in its entirety. The hotel is illuminated, highlighting its unique design and making it a beacon of light in the night sky. The video is a testament to the architectural marvels that Dubai is known for, showcasing the city's commitment to innovation and luxury."
+    ],
+    "video_ids": [
+        "GnK4WiRlj9w_59_75to209"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Burj Al Arab (hotel)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the Burj Al Arab hotel, which is accurately depicted with its iconic sail-shaped structure, illuminated at night, and situated on the beach with ocean waves. The text overlay also correctly identifies it as 'THIS HOTEL COST 1 BILLION DOLLARS TO CONSTRUCT!', aligning with known facts about the building. The visual representation matches the description of the object without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a Burj Al Arab hotel, which is characterized depicted with its distinctive sail-like design. golden at night. and situated on a water. a views in The architecturalural in confirms identifies the as theBur ISEL ISS10ION DOLLARS TO BUILDSTRUCT.' whiching with the facts about the hotel's The video elements is the description of the Bur without any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3fa98d59889f57acf10352aa4eb0845af95143f2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24d8d3e481f34169f9c9f5b184c7c3c0e06c564f21dedac3b1a8359746cc61b8
+size 210136
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..5ca614dd750c3619bc9214a3d819634047fac7fc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a drone shot of a tropical island, showcasing its natural beauty and the activities taking place on the island. The island is surrounded by clear blue waters, and there are several boats visible in the water, indicating that the island is a popular destination for boating and water sports. The island itself is covered in lush greenery, with a few buildings scattered around, suggesting that it is inhabited. The video captures the island from a high angle, providing a comprehensive view of the island and its surroundings. The style of the video is realistic, with a focus on the natural beauty of the island and the activities taking place there."
+    ],
+    "video_ids": [
+        "5_qi8yscSuU_37_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Several boats in the water, a few buildings on the island.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows several boats in the water, particularly near the shoreline of the island. Additionally, there are visible buildings along the coast of the island, which matches the description. The presence of other elements like distant islands and vegetation does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows several boats scattered the water around scattered near the island of the island. There, there are a buildings on the coastline of the island, which align the description. The presence of the elements like the islands and the does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e4510c924f44b99f6c772e3ccd883634244fc7bf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b05d1094448731cf02699d6f65bfb0cc5b829632cd48c985bd9ac443e7fab91b
+size 264167
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..0a6be8613641ca4fd29fe4b3cff9f2922250bc4d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a rocket engine, focusing on its intricate details and components. The engine is composed of multiple layers of metal, each with a unique pattern and texture. The camera captures the engine from different angles, highlighting its complexity and the precision of its design. The engine is set against a backdrop of a dark room, which contrasts with the metallic sheen of the engine, making it stand out. The video is a study in engineering and design, showcasing the craftsmanship and technology that goes into creating such a powerful machine."
+    ],
+    "video_ids": [
+        "xuVrmplbxLU_12_0to144"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A rocket engine? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows large, cylindrical rocket engines with visible internal structures and exhaust nozzles, which are characteristic features of rocket engines. The setting appears to be an indoor exhibit or museum, which is a plausible location for displaying such engines. The visual details align with the description of a rocket engine.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts a, metallic components engine with multiple internal components and components nozzles, which are characteristic of of rocket engines. The metallic and to be a engine environment or a, which is a plausible environment for showcasing such engines. The presence details, with the typical of a rocket engine,}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..718dc1f6726f5e276a82c26edec5665045a93534
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c12cdb11996b1eb5602a3758bd2bb187c628fad4cf9bc008e6b60b10cba5471
+size 282822
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..60748e3b9c3422a9f32564e6204b242bde0798c4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a person working on a motorcycle. The person is seen in the middle of the video, holding a tool and adjusting a part of the motorcycle. The motorcycle is blue and has a large engine. The person is wearing a black shirt and is focused on the task at hand. The motorcycle is parked on a concrete surface, and there are no other people or objects in the frame. The style of the video is a simple, straightforward documentation of the person working on the motorcycle."
+    ],
+    "video_ids": [
+        "Nd7RccLsMxo_32_0to156"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person and a motorcycle? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a person's hands working on a motorcycle engine, which fulfills the 'Object(s)' condition of a person and a motorcycle. The focus is on the mechanical work being performed, and no elements contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a person interacting hands interacting on a motorcycle.. which fulfills the 'Object(s)' condition of having person and a motorcycle. The presence is on the interaction interaction being done on and the additional contradict this core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5953fc994e2baab98f3f981f1f2663fde19e2d05
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e531029406cadd93d22c43fe69e6f88ed81af4bbd87a192c373580514a032139
+size 217785
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..0279bdc3ef9d3c2ce987bd8db709c57087047799
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment on a football field. The main focus is a player wearing a red and white uniform with the number 11 prominently displayed. He is wearing a red helmet and a white face mask, and his mouth is open as if he is shouting or calling out a play. The player is in motion, running towards the camera, which adds a sense of urgency and excitement to the scene. The background is filled with other players, some of whom are wearing helmets and pads, indicating that they are also part of the game. The overall style of the video is action-packed and energetic, capturing the intensity and passion of the sport."
+    ],
+    "video_ids": [
+        "3f_KEZLgG7s_7_149to270"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player wearing a red and white uniform with the number 11, other players in helmets and pads.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a red and white uniform with the number 11, wearing a helmet and pads, consistent with the description. Other players in helmets and pads are visible in the blurred background, fulfilling the 'other players' condition. There are no elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player wearing a red and white uniform with the number 11, which a helmet and pads, which with the description. The players in similar and pads are also in the background background, fulfilling the 'other players' condition. The are no elements in contradict the given description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0cef6cc0733d1245e15dfd936069d2199ccd4cb2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2edb9fc5f52fe8b7b909a932d9a20d4e3055b6c3e1847dbe1f99ce549aae3ed4
+size 196650
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..ae85d926dc55bc1320e5c553236f9dda9ae40961
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young girl and a large brown bear are seated at a dining table, celebrating a birthday. The girl, wearing a pink dress, is seated on a chair, while the bear, wearing a party hat, is standing. The table is adorned with colorful plates, cups, and bottles, and a cake with candles. The bear is leaning over the table, looking at the cake with interest. The scene is set in a room with pink walls and framed pictures on the wall. The overall style of the video is whimsical and playful, with a focus on the interaction between the girl and the bear."
+    ],
+    "video_ids": [
+        "0xIryHTVzAs_197_63to227"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young girl in a pink dress, a large brown bear wearing a party hat, colorful plates, cups, and bottles, and a cake with candles.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a young girl in a pink dress, a large brown bear wearing a party hat, and a festive table setting with colorful plates, cups, bottles, and a cake with candles. All elements described are clearly visible and consistent with the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a young girl in a pink dress, a large brown bear wearing a party hat, colorful a cake table setting with colorful plates, cups, and, and a cake with candles. The elements match in present visible and match with the scene depicted}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..dfd3b4dfa7ea203e14fba7ae1bf826e27868e0bf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:efbe4ab5df8a4e6b09f5c19f677eb0a72b479db787b2297a8f701a69dcd9e682
+size 252272
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..9fef3a98cef796c1ec996529f5bc875d7c5f7664
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a baseball player in action, showcasing his skills and the intensity of the game. The player, dressed in a white and blue uniform with the Chicago Cubs logo, is seen in three different frames. In the first frame, he is seen holding a black baseball bat, ready to swing. In the second frame, he is in the midst of swinging the bat, demonstrating his power and precision. In the third frame, he is seen running towards the base, his body leaning forward in anticipation. The background of the video reveals a crowd of spectators, their attention riveted on the player, adding to the excitement of the moment. The video is a dynamic representation of a baseball game, capturing the player's movements and the atmosphere of the event."
+    ],
+    "video_ids": [
+        "2TZLpbLWEDs_37_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Baseball player, black baseball bat, spectators? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a baseball player holding a black baseball bat, wearing a Cubs uniform, and standing near a catcher. In the background, spectators are visible in the stands, fulfilling all elements of the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a baseball player in a black baseball bat, and a uniform uniform, and standing in a dug. The the background, there are visible, the stands, watching the the of the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7488b4fe9283c48bf5f36870be4b042aabd6f5e8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:01a312b4e1e87b9808f345ed64939e20019b7dd04783401dc05b78b104798982
+size 195224
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..c864409c20820fcb7d9b4def7b565b64486d0af1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a red Honda car parked on a grassy field with trees in the background. The car is stationary and the focus is on the front of the vehicle. The style of the video is a straightforward, real-life depiction with no additional effects or animation. The car is the main subject of the video, and the surrounding environment provides context for the setting. The video likely captures the car's design and color, as well as the natural beauty of the outdoor location."
+    ],
+    "video_ids": [
+        "G4ehT5LPsv4_17_148to283"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red Honda car (front view)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a red Honda car from a front view, matching the specified condition. The car is clearly visible, and the surrounding elements (grass, trees, sky) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a red Honda car from a front view, which the description '. The car is clearly visible and and the front environment,fall, leaves, and) do not contradict the description description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..22bdee013c65a5d3ffb7484827e9d4163b2850b1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1385-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0485f970abb700998117d51a3038a4f79eb5c314887347546a152d4c0a53d7f6
+size 162227
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..03b5236b298dc32349cc646c41f5bc32de0ca2d7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young man with short brown hair is captured in three frames. He is wearing a white t-shirt and a dark gray hoodie. In the first frame, he is looking directly at the camera with a slight smile on his face. In the second frame, he is looking to the left with a thoughtful expression. In the third frame, he is looking back at the camera with a serious expression. The background is a plain gray wall, providing a neutral backdrop that allows the viewer to focus on the man's expressions and attire. The video has a simple and straightforward style, with no additional objects or actions to describe."
+    ],
+    "video_ids": [
+        "T9LHUFwRIK0_69_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man with short brown hair, wearing a white t-shirt and a dark gray hoodie.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man with short brown hair, wearing a white t-shirt under a dark gray hoodie, which matches the description. The background is plain and does not introduce conflicting elements. The subject's attire and appearance are consistent with the specified 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man with short brown hair, wearing a white t-shirt and a dark gray hoodie. which matches the description provided The individual is a, does not distract any elements.\"\n The video's appearance and hairstyle are consistent with the given 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4e7f39c75cb4ce0d71ceb532fdc3bebf1991a542
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9581be920cef60932af74ca500fd1e3e8bca0e252d366ba5dad306e9cfe19676
+size 109434
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..452897f4cd29c88845fd4a0c37cce2eb62a2bd6a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a chef is meticulously decorating a two-tiered cake with white frosting and green fondant. The chef is using a piping bag and a small round piping tip to create intricate designs on the cake. The chef's hands are steady and precise as they work on the cake, which is placed on a white cake board. The background is blurred, but it appears to be a kitchen setting. The overall style of the video is a close-up, slow-motion shot that captures the details of the cake decorating process."
+    ],
+    "video_ids": [
+        "rZNxBWsAlt0_15_23to217"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chef, two-tiered cake, piping bag, small round piping tip? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chef in a white uniform using a piping bag with a small round tip to decorate a two-tiered cake. The cake has a mint green bottom tier and a white top tier, and the chef is piping white frosting along the edge of the bottom tier. All specified objects are present and correctly depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person decorating the kitchen uniform decorating a piping bag with a small round tip to decorate a two-tiered cake with The cake is white white green and tier and a white top tier, with the chef is adding green frosting onto the edges of the top tier. The the elements are present and the used in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..591363310eacf53054c724f2925fc90ebbb9164c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5cf07cbcd9fd51aa48d39ef1909824e7b5af1e7178198d1ac0d88898232bc2f
+size 109401
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..08f0d1e750b440a5d8bd0d259d2db3dc0c8537e3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases the interior of a Mercedes-Benz car, focusing on the driver's side. The style of the video is sleek and modern, with a focus on the luxurious design and advanced technology of the vehicle. The first frame captures the driver's view, with the steering wheel prominently displayed in the center. The second frame provides a closer look at the steering wheel, highlighting the Mercedes-Benz logo and the various buttons and controls. The third frame offers a wider view of the dashboard and center console, showcasing the car's infotainment system and other features. The video emphasizes the car's high-quality materials and attention to detail, as well as its advanced technology and user-friendly design."
+    ],
+    "video_ids": [
+        "Xto8xXrYmAU_76_87to251"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, Mercedes-Benz logo, buttons and controls, dashboard, center console, infotainment system? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly displays all the specified elements: a prominent steering wheel with the Mercedes-Benz logo at its center, multiple buttons and controls on the steering wheel and dashboard, a detailed dashboard with analog gauges and a digital display, a center console with controls and storage, and an infotainment system with a touchscreen interface. All elements are accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a the elements elements: a steering steering wheel with a Mercedes-Benz logo, the center, various buttons and controls on the dashboard wheel, dashboard, a dashboard dashboard with various anduges and digital digital display, a center console with various, a, and an infotainment system with a touchscreen interface. The these are consistent represented and any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9c5cd9e7136e8e580e82b4cb09c0679c60fc97c6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aabc59afa7ca1864780081be666e46ef8643643fd5e6e49314b939a4521ea7b3
+size 179405
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..bb927a7dddbcc17754688b6e48e0481031358d59
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows two men, both wearing glasses, engaged in a conversation while standing outdoors. The man on the left is holding a small object, possibly a piece of food, and appears to be showing it to the man on the right. The man on the right is looking at the object with interest. The setting is a city street with buildings in the background, and the sky is partly cloudy. The style of the video is casual and candid, capturing a moment of interaction between the two men."
+    ],
+    "video_ids": [
+        "Ev4HypgLFoc_26_18to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men wearing glasses, one holding a small object (possibly food).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men wearing glasses. One of them is holding a small object that appears to be food (a piece of fried chicken). The description matches the core elements of the video without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows two men wearing glasses. One of them is holding a small object, appears to be food,possibly piece of bread food). The setting and the core elements of the video, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5eacccbcd6f7d2943adf30fb80158031ddbad873
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:229981699727a63984a71c1e62d4d78cb85bf7507b813486cf4f4eedeb2917a4
+size 185353
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..38205df5ef31d3fac09a99987db1efdc4863ece5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a pair of blue jeans hanging on a hanger. The jeans are displayed in a store, with the hanger and other clothes in the background. The jeans are well-made, featuring a button and a zipper fly. The denim fabric has a nice texture, and the jeans appear to be of a medium wash. The video is likely a product display or advertisement, showcasing the jeans' design and quality."
+    ],
+    "video_ids": [
+        "QzW4Gk6YEow_41_0to169"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A pair of blue jeans on a hanger.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a pair of blue jeans hanging on a metal hanger, which matches the core description. The focus is on the waistband, button, and pockets of the jeans, and while there are other hangers and jeans in the background, they do not contradict the primary object being shown.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a pair of blue jeans hanging on a h hanger. which align the description description of The jeans is on the jeansband and pockets, and pockets of the jeans, and there there are some elementsangers and a in the background, they do not detr the main subject being described.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..012d28eb3ff25eaa521642b90ba722d0da374f99
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:802d8753c83db2cea854a6bf536affe509cb109b5bd3d444192cd938ffe7c0ea
+size 241443
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..f462f2c2ea5d95773c127c3aad4810b65299b8f3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two men are in a motorcycle gear store. The man on the left, wearing a gray t-shirt with the number 38 on it, is pointing at the back of the man on the right, who is wearing a black jacket. The man on the right is facing away from the camera, and the man on the left is gesturing towards him. The store is filled with motorcycle gear, including helmets and jackets, which are displayed in the background. The lighting in the store is bright, and the overall atmosphere is casual and friendly."
+    ],
+    "video_ids": [
+        "3PeRGsNkqd8_22_0to144"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two men, one in a gray t-shirt with the number 38, and one in a black jacket.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two men: one wearing a gray t-shirt with the number 38 on it, and the other wearing a black jacket. The description accurately matches the visual content of the video, with no contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows shows two men, one wearing a gray t-shirt with the number 32 on the, and the other in a black jacket. The description is matches the objects elements of the video, with no contradictions or}<|im_end|>\n addCriterion",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0ac011fc3ac1ec05cce086397567f79d9b03e406
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9311a2b8b81795214594a9755e3dbd0ea5edcd2f0594c28813a2ac2e2993ec1f
+size 150400
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..2946f95c314c93c69cce698fce202df2bea76934
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a man in a red sweater and beige jacket standing next to a white sports car with a black interior. He is pointing at the car, possibly explaining its features or discussing its design. The car is parked in a showroom, and there are other cars and people in the background, suggesting that this is a car show or exhibition. The man appears to be knowledgeable about the car, and his enthusiasm is evident in his gesture. The setting is indoors, and the lighting is bright, highlighting the car's sleek design and shiny exterior. The overall style of the video is informative and engaging, with a focus on the car and the man's commentary."
+    ],
+    "video_ids": [
+        "CGwK821taPI_21_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Man in a red sweater and beige jacket, white sports car with a black interior? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a red sweater and beige jacket standing next to a white sports car with a black interior. The core elements described are accurately represented in the video, even though there are additional background elements like other people and posters.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man wearing a red sweater and a jacket, next to a white sports car with a black interior. The man elements of in present represented in the video, and though there are additional people elements like other cars and cars,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..55f69eb6ab8fb87901396e0a70dfee07dbedd859
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30059ff216fbcb039089e93105758f5cc07511490fa927d3c2113d6e78fd70d0
+size 213837
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..4d9e95ef49b310c409adcaabd9374389ad06be46
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man driving a vintage car. He is dressed in a suit and tie, suggesting a formal or professional setting. The car is a classic model, with a distinctive design that includes a large steering wheel and a prominent dashboard. The man is focused on the road ahead, indicating that he is in control of the vehicle. The car is moving along a road that is lined with trees and a stadium, suggesting that the location is a city or town. The overall style of the video is realistic, with a focus on the man and his actions. The video does not contain any special effects or animation."
+    ],
+    "video_ids": [
+        "M_SAOU0hCho_67_61to208"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a vintage car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man dressed in a suit sitting inside a vintage car, which matches the 'Object(s)' condition. The car's classic interior design and the man's position as the driver are consistent with the description. Additional elements like the racetrack and stadium in the background do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a man driving in a suit driving in a vintage car. which matches the descriptionObject(s)' condition. The car appears design design and and the man's attire behind the driver are consistent with the description. There elements like the roadetrack and trees in the background do not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fca078a7f0af7bd475c9d7cd97f1dd4b9946e055
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1386-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d783e7a85df84d56565114f4e7b9efabfb23f50623326d6d4cc628bd29a74710
+size 332426
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..85aafd2873ef1d0ab3038bfc43d8f83df88a1572
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a beautiful floral arrangement being created. The arrangement features a variety of flowers, including pink and white roses, purple and blue flowers, and green foliage. The flowers are arranged in a circular pattern, with the roses being the centerpiece. The arrangement is hung from a chain, suggesting that it is being used as a decoration or centerpiece. The style of the video is artistic and creative, with a focus on the beauty of the flowers and the skill involved in their arrangement. The video likely captures the process of creating the arrangement, from start to finish."
+    ],
+    "video_ids": [
+        "7NsZIMOpmxc_21_0to143"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Pink and white roses, purple and blue flowers, green foliage? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a floral arrangement that includes pink and white roses, purple and blue flowers (such as lavender sprigs), and green foliage, matching the described elements. The arrangement is displayed in a hanging basket, and all specified flowers and foliage are clearly visible.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a wre wre that includes pink and white roses, purple and blue flowers,which as hydr andigs), and green foliage. which the description elements. The arrangement is presented in a circular wre, which the the types and foliage are present visible.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..73fb10c2154adf4b74bb0e57fa991f361e65bd46
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1920cadf45808fc732ac9bc57a84476b74dce193b4e323ae2212c4857c2c2adb
+size 204200
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..d363559bbeb850f222c817503fc0f806f7e46819
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. A player, wearing a white and blue uniform with the number 21, is in the process of catching a football. He is wearing a helmet and gloves, and his body is in motion, suggesting a high-speed play. The player is the main focus of the video, and his actions are the central element of the narrative. The background is filled with other players, indicating that this is a team sport. The style of the video is realistic, capturing the intensity and excitement of the game. The camera angle is dynamic, following the player as he moves, which adds to the sense of action and movement. The colors are vibrant, with the white and blue of the uniform standing out against the green of the field. The overall impression is one of excitement and anticipation, as viewers wait to see if the player will successfully catch the ball."
+    ],
+    "video_ids": [
+        "OWReKlebHU4_4_106to255"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A player in a white and blue uniform with number 21, other players in the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a white and blue uniform with the number 21, consistent with the description. Although there are other people in the background, they do not contradict the core description and are acceptable as additional elements. The player is clearly identifiable as the main subject, and the uniform and number match the specified criteria.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a player in a white and blue uniform with the number 21, which with the description. The there are no players in the background, they are not appear the core description and are likely as additional elements.\"\n The player's also the by the main subject, and the video and number match the description details.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0749970f0f3e461df08d3694982a85a08e79f583
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3183b8143fa9cdc4d8195ba719302465d423076bf03f65a73f66acdf04f9eaa7
+size 251655
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..692fbabcab9f738900b8209dcddde1bf7e7e1de5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video depicts a meticulously crafted nativity scene, set in a miniature village. The scene is composed of various elements, including a stable, a manger, and a group of figurines representing the nativity story. The stable, constructed from white blocks, houses a manger filled with hay, and is adorned with a thatched roof. The figurines, crafted from a variety of materials, are arranged within the stable, depicting the birth of Jesus. The village surrounding the stable is composed of white buildings, each intricately detailed, and is populated with miniature figures. The scene is set against a backdrop of a blue sky, and is illuminated by a single light source, casting a warm glow on the scene. The overall style of the video is realistic, with attention to detail evident in the craftsmanship of the miniature elements."
+    ],
+    "video_ids": [
+        "bl7naS5ILlM_13_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Stable, manger, hay, figurines (representing the nativity story), white buildings, miniature figures? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a stable (manger), hay, figurines representing the nativity story, white buildings, and miniature figures. The camera pans slowly across the scene, allowing viewers to observe all these elements in detail without any contradictions to the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting a stable withaanger) hay, figurines representing the nativity story, white buildings, and miniature figures. The scene captures across across a scene, highlighting viewers to see these these elements. detail. any contradictions.\"\n the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1078db367730ee66ff3f3c9ea8266b51384acfc0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d80f866bf4cf7971a0b23b1ae584f90c4bf0faed47a4f6fe0e7e3dd24c2b6866
+size 140269
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..be28111f57bb193de212cb80a3ad85a48e4df7ee
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up of a glass being filled with whipped cream. The glass is placed on a wooden table. In the first frame, the glass is empty and the whipped cream is being poured from a bowl. In the second frame, the glass is partially filled with whipped cream. In the third frame, the glass is completely filled with whipped cream. The style of the video is a simple, straightforward demonstration of a food preparation technique. The focus is on the glass and the whipped cream, with no additional elements or distractions. The video does not contain any text or narration. The overall impression is one of a quick, easy-to-follow recipe or cooking tutorial."
+    ],
+    "video_ids": [
+        "ZyjA9ZmpPbU_19_0to116"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A glass, a bowl of whipped cream? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a glass with Oreo cookies at the bottom and a spoon adding whipped cream into it. A bowl containing more whipped cream is also visible in the background, fulfilling the 'Object(s)' condition as described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a glass and areo cookies inside the bottom, a bowl, whipped cream on the. The bowl of the whipped cream is also visible, the background, which the 'Object(s)' condition as described.\"\n}<|im_end|>\n addCriterion",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ae43cf2e0bee7cca6a4a49cc8101a5076d47e355
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:948fd368e91f366fd40d1040b1bd10a3376dc0eab5aff346e75828cd7a5aa412
+size 128064
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..4ce24c16aa1d2877a709be52e8e7548d2dc01534
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and mustache, wearing a red shirt and a black tie. He is standing in front of a blue background with white circles. The man appears to be speaking or presenting, as he is looking directly at the camera. The style of the video is a straightforward interview or news segment, with a focus on the man and his message. The background is simple and uncluttered, allowing the viewer to concentrate on the man and his words. The lighting is bright and even, highlighting the man's features and the colors of his clothing. The overall impression is of a professional and polished presentation."
+    ],
+    "video_ids": [
+        "L7oklmbtxoY_5_0to199"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and mustache, wearing a red shirt and a black tie.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and mustache, wearing a red shirt. Although the tie is not clearly visible or appears to be a patterned shirt collar, the core description of a man with a beard and mustache in a red shirt is largely accurate. The presence of a tie is not clearly confirmed, so it does not contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and mustache, wearing a red shirt and The the tie is not clearly visible, mentioned to be a solided tie,, the description elements of the man with a beard, mustache in a red shirt is met fulfilled. The presence of a pattern or not explicitly defined, but it is not contradict the description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..838e3404623f0dc5226a51c67f3768da6517d38d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:02f8c63a437bcad4ce7ff1d3af688506258097cf4f64b086b51f46da8a0823ee
+size 141788
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..47bcee78bdb24995a22986f33a0cb943e73dff8c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a delightful seafood meal being enjoyed on a wooden table. The main focus is on two white plates filled with fresh oysters and shrimp. The oysters are open, revealing their creamy interiors, while the shrimp are cooked to perfection, their pinkish hue indicating they are ready to be savored.   In the background, a cardboard box filled with crispy french fries adds a comforting element to the meal. The fries are golden brown, suggesting they are freshly cooked and crispy.   Adding a touch of freshness to the meal, there are lemon wedges scattered around the table. Their bright yellow color contrasts beautifully with the other food items, and their presence suggests they are ready to be squeezed over the oysters and shrimp for added flavor.   Overall, the video presents a mouth-watering seafood meal, with the oysters and shrimp as the stars of the show, accompanied by a side of crispy fries and a hint of freshness from the lemon wedges. The wooden table provides a rustic backdrop, enhancing the overall ambiance of the meal."
+    ],
+    "video_ids": [
+        "NkaKkcbadhQ_90_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two white plates, fresh oysters, shrimp, a cardboard box with crispy french fries, lemon wedges? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows two white plates with fresh oysters and shrimp, a cardboard box containing crispy french fries and lemon wedges. All specified elements are present and accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful two white plates with fresh oysters and shrimp. a cardboard box with crispy french fries, lemon wedges. The the elements are present and match depicted, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..adce6c35614758703ce80f46dc0a6a66b4142cf5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4547916aabf3f6e3fc90b028f784548024c3693076dfcb039cbc2cd24b90266a
+size 103569
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..e987d8a81b84f69f30e579c6241c3a1c4e6ccc9b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a group of people walking along a narrow, winding path that hugs the side of a steep cliff. The cliff is made of red rock and is quite steep, with the path appearing to be carved into the side of the cliff. The path is narrow and winding, with the cliff on one side and a steep drop-off on the other. The group of people is spread out along the path, some walking ahead and others following behind. The path is lined with a stone wall on one side, providing a sense of security for the walkers. The video is shot from a high angle, looking down on the path and the people walking along it. The overall style of the video is dramatic and adventurous, capturing the thrill of walking along a cliff-side path."
+    ],
+    "video_ids": [
+        "8WLi9l-qr3g_36_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A group of people? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a group of people standing along a narrow path on the side of a cliff. They are visible throughout the frames, confirming the presence of a group of people as described in the condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts shows a group of people walking on a narrow pathway that a side of a steep. The are positioned and the frames, and the presence of a group of people as described.\"\n the condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7882bd7ab0f9b4cf05cfef0940f7fb2541696ba9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:33b3f4d1d7009bcc93751172f9779642c5e2f15fa507f4b5c0ebf193c776ed94
+size 265144
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..d7069f7e911e7fc644a83f3956d6f477346ebfda
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up portrait of a young woman with long brown hair and blue eyes. She is smiling and looking directly at the camera. The lighting is soft and warm, highlighting her features and creating a gentle glow on her skin. The background is blurred, focusing the viewer's attention on her face. The style of the video is a simple, elegant portrait with a focus on the subject's expression and the soft, warm lighting."
+    ],
+    "video_ids": [
+        "D4f2-ToVUcg_68_0to104"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young woman with long brown hair and blue eyes, smiling and looking directly at the camera.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young woman with long brown hair and blue eyes, who is smiling and looking directly at the camera. The description matches the visual content accurately, with no conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a young woman with long brown hair and blue eyes, smiling is smiling and looking directly at the camera. The description matches the visual elements of, with no additional elements present}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f6d7f3c9ece92a21c5f8e03a1b0b49bf947415de
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1387-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f332e573bc7d95547d5f03e3876fcbba7c4d5e7297d54dd18a2ab149b4c7bff5
+size 83492
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..9a4ba1746fb17f64f3bff3c476515b91208e66a6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the growth and transformation of a plant over time. The first frame shows a small, green leaf sprouting from a branch. In the second frame, the leaf has grown larger and is now a vibrant shade of green. By the third frame, the leaf has reached its full size and is a deep, rich green color. The plant's growth is evident in the progression from the first to the third frame, showcasing the plant's development and vitality. The style of the video is a close-up, time-lapse shot that focuses on the plant's growth, providing a detailed view of the plant's transformation over time."
+    ],
+    "video_ids": [
+        "tFu8U9ve9XI_29_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A leaf? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features leaves, specifically those of a plant with glossy, green foliage. The focus remains on the leaves throughout the clip, fulfilling the 'Object(s): A leaf' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a in showing showing of a plant, a green green leaves. The sequence is on the leaves throughout the frames, showing the 'Object(s)' A leaf' condition.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..6a361d26e8717f7326d8158eefc2a9d901fbe715
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4811e12dba666f401c42c964e111cfa6a0992d6c7e384defbdf1cf667db5c046
+size 87475
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..c0c7d36a5ea34f3214f936b61176077875b59fc6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a chef is seen preparing a meal in a kitchen. The chef, dressed in a blue apron, is standing at a wooden table, which is covered with various cooking utensils and ingredients. The chef is using a knife to chop ingredients, and a spoon to mix them. The kitchen is well-equipped, with a refrigerator and a shelf filled with various bottles and containers. The chef's actions suggest that they are in the process of cooking a meal, possibly a dessert, given the presence of a cake pan on the table. The overall style of the video is realistic, capturing the everyday activities of a chef in a kitchen."
+    ],
+    "video_ids": [
+        "lrLJtId21To_4_108to255"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chef, wooden table, cooking utensils, ingredients, knife, spoon, cake pan, refrigerator, bottles and containers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a chef in a blue uniform, a wooden table, various cooking utensils (including a knife and spoon), ingredients (such as a bowl of crumbs and jars), a cake pan, a refrigerator in the background, and bottles and containers on shelves. All specified elements are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing a chef in a kitchen apr, a wooden table with various cooking utensils,knife a knife and a), ingredients (such as a lemon of flour and a of a cake pan, a refrigerator, the background, and bottles and containers on the. The these objects are present and contribute with the scene.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..cc331e0c7a17b89c0a70d7555051fc5cc8039ae2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe3ac7101a88015c863a74c855043e6d438b31849747f3dd39249010796d7116
+size 169289
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..a18c822a7332a9cdb2c7a18de1843fa58ef268ac
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a vibrant and lush depiction of a tropical jungle scene. It features a waterfall cascading down a cliff face, surrounded by a variety of exotic plants and flowers. The waterfall is the central focus of the video, with its water flowing down the cliff and creating a misty effect. The plants and flowers are in full bloom, adding a burst of color to the scene. The video is shot in a way that captures the beauty and tranquility of the jungle, with a focus on the waterfall and the surrounding flora. The overall style of the video is realistic, with a high level of detail and a focus on the natural beauty of the scene."
+    ],
+    "video_ids": [
+        "2TdhmjbQS2c_14_0to104"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Waterfall, exotic plants, and flowers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a waterfall cascading down a rocky cliff, surrounded by lush greenery. Exotic plants, including large succulent-like flora and ferns, are prominently featured in the foreground and midground. While the video does not explicitly show flowers, the presence of colorful, small blooms near the base of the waterfall can be interpreted as flowers, fulfilling the condition. The scene is rich in natural elements consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a waterfall cascading down a rocky cliff, surrounded by lush,ery and Theotic plants and including large leavesulent-like leaves, vibrants, are visible featured. the foreground. backgroundground. Vibr the plants is not explicitly show flowers, the presence of vibrant bloss vibrant flowers in the base of the waterfall suggests be considered as flowers, fulfilling the ' of The overall is vibrant in detail elements that with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..752ac4884e372deaa1f6712784e09a17c688d6c9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f1b812d129be280ef0546e776776093baf4ded8d090f585488962529386c8a29
+size 299164
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..dc6e6e048b856199eb02e245655a4f16092cad1b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is standing in front of a blue car, wearing a blue t-shirt with the words \"Ball is Back\" printed on it. He has a beard and is looking surprised. The car is parked on a street with trees in the background. The man's expression and the text on his shirt suggest that he is excited about something related to basketball. The overall style of the video is casual and candid, capturing a moment of surprise and excitement."
+    ],
+    "video_ids": [
+        "4gmZH7ls6nU_25_39to201"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a blue car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man standing in front of a blue car, which matches the 'Object(s)' condition. The man is wearing a blue t-shirt and is positioned in front of a blue vehicle, with no elements contradicting this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man standing in front of a blue car. which ful the 'Object(s)' condition. The man is wearing a blue shirt-shirt, the positioned in front of a blue vehicle, which trees additional ining the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fb611d9166375df7ad76afe5b61c5cb65338a168
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4794f2b1d6de0f3b3da3df1805711d9aa67c1c868b06da9410947955f0172e90
+size 124775
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..340f9efb1a69b19b8ef6a769f00516b713e0add9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the growth of two corn plants in a garden. The first frame shows the plants just beginning to sprout from the soil. The second frame shows the plants growing taller, with their green leaves beginning to unfurl. The third frame shows the plants fully grown, with their leaves fully extended and their stalks reaching upwards. The garden bed is filled with rich, dark soil, and the plants are surrounded by a wire fence. The plants are the main focus of the video, with their growth and development being the central theme. The style of the video is a time-lapse, capturing the plants' growth over time."
+    ],
+    "video_ids": [
+        "2_1BPW6hrRY_52_0to134"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two corn plants? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows two young corn plants emerging from the soil, which matches the 'Object(s)' condition of 'Two corn plants'. The presence of other elements like soil, garden fencing, and background vegetation does not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows the corn corn plants growing from the soil, which align the descriptionObject(s)' condition. theTwo corn plants'. The progression of additional plants like the and a background, and additional green does not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8fb45dcf91dd3dfb46fecd2fb1453146c66bdf06
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59d29bc9d7e7e7f4e4b8c4d43eb5d6cf09ad57cc516d8274304a649fa30ec604
+size 210038
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..ded765fd8ca65896d4c3c65edd9b26a371007686
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a football game. The main focus is on two players, one in a blue and yellow uniform and the other in a white and green uniform. The player in the blue and yellow uniform is in the process of tackling the player in the white and green uniform, who is holding a football. The action is intense, with the player in the blue and yellow uniform lunging forward, while the player in the white and green uniform is trying to evade the tackle. The background is filled with other players, all in various stages of action, adding to the overall excitement of the scene. The video is shot from a side angle, providing a clear view of the action and the players' movements. The style of the video is realistic, capturing the intensity and excitement of the game."
+    ],
+    "video_ids": [
+        "YIxSvf1JV2o_4_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two players - one in a blue and yellow uniform and the other in a white and green uniform. The player in the white and green uniform is holding a football.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two players as described: one in a blue and yellow uniform (number 50) and another in a white and green uniform (number 12) holding a football. The visual elements align with the description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows two players in described: one in a blue and yellow uniform andthe 71) and the in a white and green uniform (number 71). holding a football. The player elements match with the description, any.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c5967e307eccd31e89efeb215d72d7114d30a28c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26b82e08b933da670e3989a7a34ca5910f0c137135ff0dcdd77beda220a432a2
+size 371333
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..ded7d8b357b0c66d537e0c035a1f9446abcea491
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a sandy environment where a small insect, likely an ant, is actively engaged in digging or moving sand particles. The scene is dominated by a textured surface of fine, light brown sand grains, which vary slightly in color and size. The insect, positioned centrally in the frame, is seen using its legs to push sand particles aside, creating a small disturbance in the sand. As the insect moves, it occasionally lifts its antennae, possibly sensing its surroundings or communicating with other insects. The background remains consistent throughout the frames, emphasizing the detailed texture of the sand and the focused activity of the insect. The BBC logo in the top left corner suggests that this footage is part of a documentary or educational program. There are no significant changes in lighting or camera angles; the focus remains steady on the insect's actions, providing a clear and detailed observation of its behavior in this natural setting."
+    ],
+    "video_ids": [
+        "e88da73f84bd9f52551a9a5788ec3617b53fe6dc70b82c08a758ee06012f361d"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small insect (likely an ant) and a textured surface of sand grains.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a small insect, consistent with an ant, interacting with a textured surface composed of sand grains. The ant is seen moving and digging into the sand, which aligns with the described condition. There are no conflicting elements that contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a small insect, likely with the ant, situated with a textured surface that of sand grains. The insect is positioned moving across navigating into the sand, which fuls with the description conditions. The are no additional elements in contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7796623eb8f010ac065e94ca6e1b98079b9ab50a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:871ea9bdaa70dc7cef83e59648cb54a87ff062fe4f928217eb0e328e49fab4b8
+size 120650
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..d88d7318bfd2b135eb6fb626222f9155c4821a47
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a split-screen comparison of a motorcycle in two different states. On the left side, the motorcycle is shown in its full form, with a sleek design and a black and silver color scheme. The motorcycle is parked on a gray surface, and the background is a plain white wall. On the right side, the motorcycle is shown in a more detailed view, with a focus on its rear wheel. The wheel is lifted off the ground, revealing the intricate details of the motorcycle's suspension and braking system. The background is the same plain white wall, but the motorcycle is now positioned on a red stand, which contrasts with the gray surface it was previously on. The text \"SO-CALLED AS THEIR PURPOSE IS\" is displayed at the bottom of the screen, suggesting a commentary on the motorcycle's design and purpose. The style of the video is informative and technical, with a focus on the motorcycle's features and design elements."
+    ],
+    "video_ids": [
+        "sgbZNtDBlfE_22_102to291"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Motorcycle (full form and rear wheel detail), red stand, gray surface, plain white wall, text 'SO-CALLED AS THEIR PURPOSE IS'? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a full-form motorcycle with detailed rear wheel, a red stand, a gray surface, a plain white wall, and the text 'SO-CALLED AS THEIR PURPOSE IS'. All specified elements are present and accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as displaying showing a motorcycle form motorcycle and a rear wheel. a red stand, a gray surface, and plain white wall, and the text 'SO-CALLED AS THEIR PURPOSE IS'. The elements elements are present and do depicted in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3c37cba5288d3d26d869e8800463c91b04d48225
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1388-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21254f6120fbedf56c897b5454215a6e8de9e5e623ade0eaf34e62317819d4d6
+size 128665
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..d39b4115c1ff5fd19c3b170adf5fffa20c16c455
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man in a red jersey is sitting on a couch with two women. He is holding a piece of paper and appears to be reading it aloud. The women are listening attentively. The man is wearing a watch on his left wrist. The couch is yellow and there is a blue pillow on it. The room has a pink curtain in the background. The video seems to be a casual, indoor setting with a focus on the interaction between the man and the women."
+    ],
+    "video_ids": [
+        "Ciqs1wBWr_A_79_0to143"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a red jersey, two women, a piece of paper, a watch, a yellow couch, a blue pillow.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in a red jersey sitting on a yellow couch with two women beside him. A watch is visible on the man's wrist, and a blue pillow is on the couch to the left. There is no piece of paper visible in the video, but since the presence of additional elements is acceptable as long as they don't contradict the core description, the video largely matches the required objects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man in a red jersey holding on a yellow couch, a women. him. The piece is visible on the man's wrist, and a piece pillow is present the couch. his right. The is a piece of paper in in the video, but the the question of an elements is acceptable as long as they do't conflict the core description, the video largely matches the given elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c9e7bd1bf3c57ae791a7374f05f3eadd8d19f735
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:909e306a7847c4259dc051b881f7849c2edfe6a2e1f22db19f64668adcbe84b5
+size 110127
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..11956d2cf5d35db23942365e2f1314c468c5c9a8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a blue jacket with various sponsor logos, sitting in a room filled with automotive memorabilia. He is bald, with a gray beard, and appears to be speaking or gesturing with his right hand. The room has a greenish hue, and the background is filled with vintage car parts, including a large wheel and a collection of license plates. The style of the video is casual and conversational, with a focus on the man and his surroundings. The lighting is soft and ambient, highlighting the textures and colors of the room and the man's attire."
+    ],
+    "video_ids": [
+        "_aL5XtA69ko_52_26to216"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a blue jacket with sponsor logos, who is bald and has a gray beard. He is speaking or gesturing with his right hand.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bald man with a gray beard wearing a blue jacket adorned with sponsor logos. He is speaking and gesturing with his right hand, which aligns with the description. The background elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man man with a gray beard wearing a blue jacket with with various logos. He is gest or gesturing with his right hand, which aligns with the description provided The background,, not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b7ff0cfd86e57daca64ca45e555c47407338fda8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f9595d165fa3a1cc7ef3d4c878b4e13a6ab3c9552b6d6ce144da7a70aad0c350
+size 184126
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..3a3a907a94455cfc9628ec8ebf0795437c4f1b60
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene scene of a bird in flight over a body of water. The bird, likely a seagull or similar species, is seen gliding gracefully above the surface, its wings spread wide against the backdrop of rippling blue water. The bird's movement is fluid and smooth, suggesting it is gliding rather than flapping vigorously. As the bird flies, it passes over a small splash on the water's surface, indicating a disturbance beneath the surface, possibly from an underwater creature or a fish\u8dc3\u51fa\u6c34\u9762. The overall atmosphere is calm and tranquil, with the gentle waves and the bird's steady flight creating a peaceful visual experience. The camera remains stationary throughout the sequence, focusing on the bird and the water below, allowing viewers to fully appreciate the natural beauty and tranquility of the scene."
+    ],
+    "video_ids": [
+        "a95ffcd4b6d0da80f70c5257e9aa27c3688e6cd8fa614d2dff9fc4ab22607294"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A bird (likely a seagull) flying over the water.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bird with dark wings and a lighter underbelly flying low over the water, consistent with the description of a seagull-like bird. The bird's motion and environment match the core condition without significant contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a bird, wings wings and a lighter bodybelly flying over over a water. which with the description of a seagull. bird. The water is wings and the align the ' description of any additional.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..de9c6f462f70c045d1e43227e7540a53e05e0dc5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:22acb487be3c81903e8c71ba5ba98ef7097b6e7a1e770277d356a4ccec8e8e67
+size 143431
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..55695c46bc41f0ebe7d6edf8dcd5443a92f3ba9e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a basketball jersey, likely a professional player, captured in three frames. The first frame shows him looking to his left with a focused expression. In the second frame, he is seen looking straight ahead, still with a serious expression. The third frame captures him looking to his right, maintaining his serious demeanor. The background is blurred, but it appears to be an indoor basketball court with spectators in the stands. The style of the video is a close-up, slow-motion shot, emphasizing the player's facial expressions and the intensity of the game."
+    ],
+    "video_ids": [
+        "6wNGqWLCUCw_57_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a basketball jersey? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a basketball jersey with the word 'JAZZ' on it, which matches the description of a man in a basketball jersey. The background, though blurred, suggests a basketball court setting, and the man's attire and posture are consistent with a basketball player.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a basketball jersey, a number 'NazzZ' visible it, which is the description of a man in a basketball jersey. The background appears although blurred, appears a basketball game or, further the man's attire and posture are consistent with someone basketball player.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7afee9665ebb3dcb50fb51d8b6a09ae7e86a892a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87ee9139c315e1d653c883cf04e193b039da8ecd86371691f12d6b527dedbbbb
+size 174988
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..8a27f94b0749ffdf81a4478b652aa74dd62fee9d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene lakeside scene during what appears to be either sunrise or sunset, as indicated by the warm hues of orange and yellow reflected on the water's surface. The sky is adorned with streaks of clouds, adding texture and depth to the tranquil atmosphere. On the left side of the frame, a dense cluster of trees stands tall, their silhouettes contrasting against the vibrant sky. In the foreground, two ducks are seen engaging in their natural behavior. One duck is positioned closer to the shore, dipping its head into the water, likely searching for food. The other duck is further out in the lake, swimming calmly across the surface. The water is calm, with gentle ripples emanating from the ducks' movements, enhancing the peaceful ambiance. The overall composition of the video emphasizes the harmony between nature and the gentle activity of the ducks, creating a soothing visual experience."
+    ],
+    "video_ids": [
+        "0995ce929f47ee681a1ef9076e00a173e46f99a032b8e4f91ee40fe8404f57db"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dense cluster of trees, two ducks? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dense cluster of trees along the left bank of the water body, and two ducks are clearly visible\u2014one near the shore and one swimming in the water. These elements match the specified 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a serene cluster of trees in the left side of a lake,, and there ducks are visible visible swimming in the center and the swimming in the water. The elements match the ' 'Object(s)' condition without}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0635140217760aebed6b20c21360f7acb0d0da26
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb8a4ba89016a698300f10c1a43ba568f30aa157c4148a4facf4d346d2cb7631
+size 166678
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..34aad5faafc21356b084334060438edb320bd72b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a black sports car driving down a street lined with yellow railing. The car is sleek and shiny, reflecting the light from the overcast sky. The street is wide and empty, with the car being the only vehicle in sight. The yellow railing adds a pop of color to the otherwise gray and gloomy day. The car's speed is evident as it moves along the street, leaving a slight blur in its wake. The overall style of the video is dynamic and energetic, capturing the thrill of driving a sports car on an open road."
+    ],
+    "video_ids": [
+        "eIgmdeYHttU_23_0to102"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A black sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a black sports car, specifically a Lamborghini Aventador, parked in front of a building with yellow railings. The car is the central focus and matches the description of a black sports car.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a black sports car driving which a sleekorghini Hurventador, which on a of a yellow. a wallsings. The car's the central object of matches the description of a black sports car.}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4d8e502aa0a734ae87cb5688be230872fc214fbd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bfd005dc6f786aa6186d2245dfc8398834a092f873eb0fe9015c954df9c984d0
+size 149312
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..e950ebfdff8f742e75865b4f2254a1ac05b4e9a8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment on a football field, featuring a group of players in white and gold uniforms. The players are seen in various positions, some walking, others standing, and a few in motion. The focus is on two players in particular, one wearing the number 71 and the other wearing the number 67. The players are wearing helmets and pads, indicating they are in the middle of a game or practice. The field is well-maintained, with clear markings and a goal post visible in the background. The atmosphere is energetic and competitive, typical of a football game."
+    ],
+    "video_ids": [
+        "qAMCrOUSZbg_41_72to211"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Group of players in white and gold uniforms, players numbered 71 and 67, helmets, and pads.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group of football players wearing white and gold uniforms, with players clearly numbered 71 and 67 visible. They are also wearing helmets and pads, matching the description. The scene is consistent with an NFL game setting, and no elements contradict the specified conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group of players players wearing white and gold uniforms, with players numbered numbered 71 and 67.. The are also wearing helmets and pads, which the description provided The presence is set with a American game setting, which the additional contradict the given conditions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..39a218f59fbfced77311f0bb32ac827c97066047
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e60f7fc93a960ee956f3e5de4839c2a5c61cc8deab635fcc03532e950bdca88
+size 248486
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..397b50293eff3a4737bc06947e1e4cba7087458e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a promotional advertisement for the \"Titan\", a luxury SUV. The car is shown in three different frames, each highlighting a different aspect of the vehicle. The first frame shows the car from a front three-quarter angle, emphasizing its large size and imposing presence. The second frame focuses on the car's interior, showcasing its luxurious features and spacious seating. The third frame provides a rear view of the car, highlighting its robust design and powerful stance. The car is painted in a vibrant shade of red, which stands out against the backdrop of a snowy mountain landscape. The video also includes text that provides information about the car's engine, transmission, and price. The overall style of the video is sleek and polished, with a focus on the car's luxury and performance."
+    ],
+    "video_ids": [
+        "88-hRTIHQFI_38_374to498"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Luxury SUV (Titan) - painted in a vibrant shade of red? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video displays a luxury SUV named 'Titan' that is painted in a vibrant shade of red. The vehicle's design, including its robust build and distinctive grille, aligns with the description of a luxury SUV. The red color is prominent and consistent throughout the frames, fulfilling the specified condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a red SUV painted 'Titan' painted is painted in a vibrant shade of red. The vehicle is design, including its grille build, sleek grille, aligns with the characteristics of a luxury SUV. The video color is vivid and vivid throughout the video, fulfilling the ' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f8b5e9f64788e4d7e30a31aa2e006bdc74d70bb9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1389-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:767cfa7cb3ebaf440499329d3b28ab7f61e68a685a2957407d2ab1b5b18eeb1c
+size 178268
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..6094ea1cb71771ca472da0a5bd4727005280da37
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit and tie, sitting in front of a microphone, likely participating in an interview or a press conference. The man is bald, wearing a dark suit with a white shirt and a patterned tie. He appears to be speaking or gesturing with his hands, suggesting he is engaged in a conversation or explaining something. The background includes a banner with the words \"DOMINANCE MANAGEMENT\" and \"FINANCE\", indicating that the event may be related to finance or business management. The style of the video is a straightforward, professional interview or press conference, with a focus on the man and his interaction with the microphone and the audience."
+    ],
+    "video_ids": [
+        "VGOqnwMCoJw_29_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, a microphone, and a banner.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man dressed in a suit and tie, speaking into a microphone, with a banner in the background featuring logos and text. All core elements described in the 'Object(s)' condition are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man in in a suit and tie, standing into a microphone. which a banner in the background. text and text. The elements elements of in the questionObject(s)' condition are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..4f87b0b2ba68d5f76636186a8921dc77d5ab9b6f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dbbdbd247b79bc6176db8eb4f13aa600f2afd3c8a819fd01a97d400c6bcc352b
+size 234309
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..e37feb1c406c8056c1216cd311c3672189307b17
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a moment of celebration on a football field. Two players, both wearing blue and white uniforms, are the main focus. The player on the left, wearing the number 24, is clapping his hands in a gesture of excitement or encouragement. His teammate, wearing the number 13, is raising his arms in triumph, possibly after scoring a touchdown. The background reveals a football field with a scoreboard displaying the number 2, indicating the score or the quarter of the game. The atmosphere is one of victory and camaraderie, as the players celebrate their success on the field."
+    ],
+    "video_ids": [
+        "LgSmmCxj3B4_9_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two players wearing blue and white uniforms (player 24 and player 13).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two football players wearing blue and white uniforms, with visible numbers 24 and 13 on their jerseys. Their attire and positioning match the description, confirming the core condition is met.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows two players players wearing blue and white uniforms, with one numbers 24 and 14. their jerseys. The actions and the on the description of fulfilling that fulfillment condition is met.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..028cd453bb0e471ea639a1e73fc88b90cbc71914
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:01d0778fa80bd8dd2beb3434ce96e28bebca9a53df066625445c81c6b21d9c42
+size 228344
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..5a4bd06e49cac7227c175e4c6703af94ab5496bd
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man with a beard and short hair, wearing a green and white polo shirt. He is seated in a studio with a yellow and green background. The man appears to be engaged in a conversation, as he is looking to his left with a slight smile on his face. The lighting in the studio is bright, highlighting the man's features and the vibrant colors of the background. The overall style of the video suggests a casual and relaxed atmosphere, possibly for a television interview or a talk show."
+    ],
+    "video_ids": [
+        "JFEn7l5OBh4_1201_31to172"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man with a beard and short hair wearing a green and white polo shirt, seated.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and short hair, wearing a green and white polo shirt, seated. The background elements (lamp, colorful wall) do not contradict the core description of the man's appearance and posture.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man with a beard and short hair wearing wearing a green and white polo shirt, seated. The background is,strip, screen wall) do not contradict the description description and the man's appearance and attire.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5623d0e1b418c718fa5a67570fbc2fad7c22100d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23a1993f21211bfd8103bc8088eca30d3ed67379203e4e261c4bb6f18ba7426a
+size 139922
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..74c7818c1008e568ea9ca4178795c5e2ac779ae8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a police officer standing next to his car. The officer is wearing sunglasses and a uniform. The car has a white exterior and is equipped with emergency lights on top. The officer is leaning on the car, looking off into the distance. The background features a clear blue sky with a few clouds. The style of the video is realistic and it appears to be a still image rather than a moving one."
+    ],
+    "video_ids": [
+        "KIiYlIeaNJk_25_0to159"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A police officer, a car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a police officer leaning against a police car with flashing lights on top, matching the core description. The officer is wearing a uniform and sunglasses, and the car is identifiable as a police vehicle with its distinctive lights and design. No elements contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a police officer standing against a police car. a lights on top. which the description description of The presence is wearing a uniform and sunglasses, and the car is a as a police vehicle with its distinctive design and design. The additional contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..64ea8f8324770866d6b74eabecd79ae9414c69b1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e666c842b847d0f1e4baaf399acfc2a56448c4a4cb52c2b1b43735a1b7a6716
+size 123934
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..bbad7fc89a3bf8b908ac73775d071eeef928c14f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a close-up view of a plate of seafood pasta, featuring linguine noodles mixed with shrimp, squid, and chunks of chicken. The dish is garnished with slices of tomato and sprinkled with herbs, giving it a vibrant and appetizing appearance. The pasta is coated in a light sauce, likely a garlic or herb-based dressing, enhancing its visual appeal. In the background, there is a glass of dark-colored beverage, possibly soda or iced tea, placed on a textured surface that resembles a tablecloth with a blue and white pattern. A fork is also visible on the left side of the frame, indicating readiness for eating. The scene remains static throughout the video, focusing on the delicious meal without any noticeable changes or movements."
+    ],
+    "video_ids": [
+        "e2a5be1d8b8b50f6ee57abbee80aa4fd393a9deb325bcdb2fafffd4cbfed87e5"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: seafood pasta (linguine noodles, shrimp, squid, chicken chunks), tomato slices, herbs, light sauce, glass of dark-colored beverage, fork? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a seafood pasta dish with linguine noodles, shrimp, squid, and chicken chunks, accompanied by tomato slices, herbs, and a light sauce. A glass of dark-colored beverage and a fork are also visible, matching the description. There are no conflicting elements in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showcasing showing a plate pasta dish with linguine noodles, shrimp, squid, and chicken chunks. along by tomato slices and herbs, and a light sauce. Additionally glass of a-colored beverage is a fork are also visible in align the description provided The are no contradictions elements in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0578cf6d6afa9fec9ee86cb0169d63ad7f736e9a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:10781ea1d4e7585c8abd92d0b00e5f9bfedecb949826734f795b7d264cc11ca5
+size 142182
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..8aea7d5d251494e91c35f64402c6b6a9c008678f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a serene beach scene with a series of small boats floating gently on the calm, blue-green water. The boats, varying in size and color, are scattered across the water, some closer to the shore while others are further out. The sandy beach is visible in the foreground, with a few palm trees dotting the landscape. In the background, a pier extends into the water, with a few more boats docked at its end. The sky above is a clear blue, with a few clouds scattered across it. The overall style of the video is peaceful and tranquil, capturing the beauty of a day at the beach."
+    ],
+    "video_ids": [
+        "ycMPXockKn0_103_0to131"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Small boats, palm trees, pier? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows small boats floating in the water, palm trees along the shoreline, and a pier extending into the water. These elements are clearly visible and align with the described conditions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful small boats scattered on the water, palm trees along the beach, and a pier extending into the sea. These elements align clearly visible and match with the description '.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..823f7cfc0453c57f881021783f325c97e0bfd1fb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:22ec7ade9862a8d8700bff10a619e7a9f2c0168c9e3547b95fe72ad2fe565e10
+size 130495
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..3c89111698ba1e5c8f5e7ee0dfb737f9b47e62ff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a scene set outdoors on a concrete surface surrounded by lush greenery and trees. A large rectangular mirror with a wooden frame is placed on the ground. In front of the mirror stands a brown and white dog, its tail raised, appearing curious or alert. The dog's reflection is clearly visible in the mirror, showing it standing in the same position. To the left of the mirror, a chicken walks around, seemingly unbothered by the dog or the mirror. Another dog, similar in color to the first, is seen in the background, standing near a small concrete structure. The dogs appear to be interacting with each other, possibly barking or communicating. The overall atmosphere is calm and natural, with the dogs and the chicken moving slightly but maintaining their positions relative to each other and the mirror. The camera remains stationary throughout the sequence, focusing on capturing the interaction between the animals and their environment."
+    ],
+    "video_ids": [
+        "10ebc7969f4dfcbcf65e99fbb3b10716c65efca0bdf54c763607f1a9804917c8"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Large rectangular mirror with a wooden frame, brown and white dog, chicken, another dog, small concrete structure? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a large rectangular mirror with a wooden frame, a brown and white dog standing in front of it, a chicken near the mirror, another dog visible in the reflection, and a small concrete structure (possibly a low wall or platform) in the background. All elements described in the condition are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a large rectangular mirror with a wooden frame, a brown and white dog, in front of it, a chicken, the dog, and dog in in the reflection, and a small concrete structure inpossibly a part wall or part) in the background. The elements match in the condition are present and do with the video content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..881d0ea11a9e0b103334a22e48f718ecf4bcdf38
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea86d53c1df1a008fbb2a9142f7a12db9c4ebfd19cafff6e6eaf9efc251e2fe1
+size 215240
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..b5732abb291d87527efbe077cbe089fb724c1c47
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the vibrant beauty of a tree adorned with bright yellow berries. The berries, round and plump, are scattered throughout the tree, creating a lush canopy of color. The tree is set against a backdrop of a blurred landscape, adding depth and context to the scene. The video is shot in a close-up style, focusing on the tree and its berries, while the background remains out of focus. This style emphasizes the tree and its berries, making them the central focus of the video. The colors in the video are bright and vivid, with the yellow of the berries standing out against the green of the tree and the blurred background. The overall effect is a visually striking representation of nature's bounty."
+    ],
+    "video_ids": [
+        "BWTRUcuUumE_6_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A tree with bright yellow berries? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a branch covered in bright yellow-orange berries, which aligns with the description of a tree with bright yellow berries. Although the branch is not part of a full tree, the berries are clearly visible and match the color description. The background elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a close with with bright yellow berries berries, which aligns with the description of a tree with bright yellow berries. The the background is not a of a full tree, the focus are clearly visible and match the color and. The background,, not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..93663b33cba5c17838c8f9c9978c564c7e081db5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-139-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea66c3564f3911f89f0afc46b2f78b7f6a08f0eb8df8c724fb8b33171d57ce37
+size 171767
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..f6421e4e9e9d03623ed5903e6fb2df2aa71621ea
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a large Icelandair airplane soaring through the sky. The airplane, painted in white with blue and yellow accents, is the main focus of the video. It's a clear day with a few clouds scattered in the sky, providing a serene backdrop for the airplane's journey. The airplane is captured in three different positions, each time flying higher and further into the distance. The first frame shows the airplane flying low, the second frame captures it mid-flight, and the third frame shows it flying high, almost out of sight. The video beautifully captures the airplane's journey from takeoff to cruising altitude."
+    ],
+    "video_ids": [
+        "oJhTO05Fi48_12_0to123"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large Icelandair airplane? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a large Icelandair airplane in flight, clearly displaying the airline's livery and branding. This fulfills the 'Object(s)' condition as described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features large large airplaneair airplane in the, which matching the airline's brandingivery and branding. The alignfills the 'Object(s)' condition as it,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0aec6c55b2a9fb709a76b131afb156ce588f10a8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5035fe1f341d29bff678461e08bc6ce1e077fe8dc102f30b291a00a34de31bca
+size 233572
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..551dc2f8588c62ea0a21dacadfb702b00b43e253
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a pink pony with a blue flower on its side, standing in front of a toy bakery. The pony has long pink hair and blue eyes. The bakery is filled with various types of donuts, including chocolate, glazed, and sprinkled. The pony appears to be looking at the donuts, possibly contemplating which one to choose. The scene is playful and colorful, with the pink pony standing out against the backdrop of the toy bakery. The donuts are arranged in a way that suggests they are ready for purchase, adding to the overall charm of the scene. The video is likely aimed at children, given the bright colors and whimsical setting."
+    ],
+    "video_ids": [
+        "LbujOUdMyHk_55_68to196"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A pink pony with long pink hair and blue eyes, a toy bakery filled with donuts? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a pink pony with long pink hair and blue eyes, which matches the description. The background includes a toy bakery filled with donuts, also matching the description. The presence of another character (a brown plush toy) does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a pink pony with long pink hair and blue eyes, which matches the description. The pony shows a toy bakery filled with donuts, also align the description. The presence of the toy ora blue pony toy) in not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0d08a3ea73b81d6e26843ff76d8998dc0a689ce5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:edad6100a557e5a433fddededcedc1b0728966ddf69d4d5d0b372df022d6142e
+size 216735
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..f2aa1e5e4d7b76d8574ac0b2ae7c8e95df7e9736
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a hiker's journey in a mountainous area. The hiker, dressed in a blue jacket and black pants, is seen setting up a green tent on a rocky hillside. The tent is pitched next to a large boulder, providing a sense of scale to the scene. The hiker's backpack is placed nearby, indicating that they are prepared for an extended stay. The background reveals a majestic mountain range, with snow-capped peaks and rocky cliffs. The sky is clear, suggesting good weather for the hike. The overall style of the video is realistic, capturing the beauty of nature and the solitude of the hiker's journey."
+    ],
+    "video_ids": [
+        "Dsfv0SF5fh0_6_0to113"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Hiker, green tent, large boulder, backpack? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a hiker standing near two tents (one green, one yellow) in a mountainous area with large boulders and patches of snow. A red backpack is visible near the green tent. All specified objects are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a hiker in next a large,one green and one blue), and a mountainous area. large boulders. a of snow. The backpack backpack is also on the h tent, The the elements ( present and the with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..addb6eb3923be6b1ae9c331643d19aae81e3fceb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:edfb3dae44490a882d405e885d6bf8f73b4c34503c561842e1401c4ccb9fbbfd
+size 324568
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..4c3935bd217a40f3a9d310f94c825dedf3d787ff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen interacting with a silver Porsche sports car. The car is parked on a road, and the man is positioned behind it, leaning on the car's rear. He is wearing a white sweater and blue jeans, and he appears to be examining the car's license plate, which reads \"RK19 FSO\". The car itself is a sleek, silver Porsche, and it features a distinctive black rear wing. The setting is a forested area, with trees visible in the background. The overall style of the video is casual and relaxed, with a focus on the man and the car."
+    ],
+    "video_ids": [
+        "OoIlk4RkS28_23_93to229"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a silver Porsche sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man crouching next to a silver Porsche sports car, specifically a 718 Boxster GTS, as described. The core elements of the request \u2014 a man and a silver Porsche sports car \u2014 are accurately represented without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a man interactingrouching next to a silver Porsche sports car. which a Porsche918 modelster,TS, as indicated. The car elements of the '\u2014 a man and a silver Porsche sports car \u2014 are present represented in any or}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e7ea9f9999c6b0bd0b376015b259dbd13ab9671b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0939510c771c8773f5d0c09c9630e84bffdeef9dfa7439d8030e542798724673
+size 218396
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..2f2d70dd1e1f13cb0b47c250c90a68aee27acda1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a soccer player in action on the field. He is wearing a blue and red jersey with the logo of the sponsor, Pmu, prominently displayed. The player is seen in three different positions: running, pointing, and celebrating. The background features a blurred soccer field, indicating the focus is on the player. The style of the video is dynamic and energetic, capturing the intensity of the game."
+    ],
+    "video_ids": [
+        "b9DQuRzAujs_6_0to130"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A soccer player in a blue and red jersey with the Pmu logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a blue and red jersey with the PMU logo prominently displayed on the front, matching the description. The player is also seen raising his hand, which is consistent with typical post-goal or celebratory actions in soccer. There are no conflicting elements that contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a blue and red jersey with the PU logo, displayed on the front. which the description provided The player is also wearing wearing one arm, which is a with the soccer-game celebrations celebratory gestures in soccer. The are no elements elements in contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b9316e36c69f6218f33a9daddfd403390f664a5e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5a720bc33cbae22f4c8a9ceb2c3e241fa15777fa44acd6c3006d255d3869107
+size 184625
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..1ea8fbeece3de2023c13d7041b407c9546242254
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a Formula 1 car in a studio setting. The car is predominantly orange and black, with the number 7 on the side. It features a blue and white front wing and a blue and white rear wing. The car is equipped with a black helmet and a black steering wheel. The car is stationary, and there is no text or other objects in the video. The style of the video is a realistic 3D rendering, providing a detailed and accurate representation of the Formula 1 car."
+    ],
+    "video_ids": [
+        "AEr2Kdj5Vsc_12_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Formula 1 car, black helmet, black steering wheel? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a Formula 1 car, which is the primary subject. The driver is wearing a black helmet, and the steering wheel is also black, matching the description. Additional elements like sponsor logos and the car's livery do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a Formula 1 car with which is orange primary object. The car is wearing a black helmet, and the steering wheel appears not black. as the description. The elements like the logos and the car's designivery are not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..390eb61992ae07e744cc4b730690b7f8902b29a4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6177dd405a7b4b7f30d9c8e9438c1d4ff78d4ed605cb0582bd6f42661224ccf
+size 95104
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..29ed95f695761be425523b7ad9981fc624ff512f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a vintage doll with curly hair, sitting in a cluttered room. The doll is dressed in a white dress and has a tear on its face. It is holding its hand to its ear, as if listening to something. The room is filled with various objects, including a chair and a pile of clothes. The overall style of the video is nostalgic and evokes a sense of longing."
+    ],
+    "video_ids": [
+        "6pOqp-CitKg_59_0to149"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A vintage doll with curly hair, dressed in a white dress, and a tear on its face. Also, a chair and a pile of clothes.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a vintage doll with curly hair, wearing a white dress, and visibly has a tear on its face. In the background, a chair and a pile of clothes (including a pink garment) are also present, matching the description. The scene is consistent with the specified elements without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a doll doll with curly hair, dressed a white dress, and there crying a tear on its face. The the background, there chair and a pile of clothes areincluding a blue item) are present present, which the description provided The presence is consistent with the given elements.\"\n any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bba781839db75566b003f26babced779f06bbe53
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9db24ce45cc250f4ea51e490e2e3d4ed26736b723af348b93e4db9ec7dc89e8e
+size 107855
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..7b75b83e8ccfeca7ed3910850a7159f607c92e9c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a woman standing at a podium, speaking into a microphone. She is wearing a blue and white patterned blouse. The podium is wooden and has a microphone attached to it. The background features a door with a \"staff only\" sign on it. The woman appears to be addressing an audience, as she is looking towards the right side of the frame. The style of the video is straightforward and documentary-like, capturing a moment of public speaking."
+    ],
+    "video_ids": [
+        "9O3kfiP38mE_7_76to209"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a podium, a microphone, a door, a 'staff only' sign? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman standing at a podium with a microphone. To her left, there is a door with a visible 'staff only' sign. All the specified objects are present and correctly identified in the scene, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a woman standing at a podium with a microphone in There the right, there is a door, a ' signstaff only' sign. The elements elements elements are present in match identified in the video.\"\n fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8070b921a900c4e314d0ee76d08f2a61270f3a97
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1390-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40889265f4451b6d4cfc34a58fde1ac8f150aa5dd207bab8f06657b281aae3e1
+size 158092
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..b9f28f242c7e458cedeb40878e3f28b99065874d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a chef in a professional kitchen, showcasing his culinary skills. The chef, dressed in a black apron and a matching black cap, stands confidently behind a wooden counter. In front of him, a white bowl filled with golden-brown fried food, possibly chicken, is placed. The chef holds a spoon, ready to serve or taste the dish. The kitchen is well-equipped with various appliances, including an oven and a sink, indicating a professional setting. The overall style of the video suggests a cooking show or a culinary tutorial, with the chef as the main subject, demonstrating his expertise in preparing delicious meals."
+    ],
+    "video_ids": [
+        "MCpJXQy0bj8_14_55to271"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chef, white bowl with fried food, spoon? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a chef standing in a kitchen, a white bowl filled with fried food (appearing to be fritters or falafel), and a spoon placed in a small compartment of the bowl. These elements match the 'Object(s)' condition described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a chef in behind a kitchen, wearing white bowl filled with fried food,likelyaring to be donritters or similarafel), and a spoon being in the bowl dish of the bowl. The elements match the descriptionChef(s)' condition provided in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..31d728e4d3751e22afa05fd8e2a1e3b8ef12503a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:05708c555cc1ef2257a67242112ce4ca76c4eefcb0cf9412404533466686fdc9
+size 103465
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..a28031222afa476516f773fd617f0aa0a0210269
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a man in a suit and tie, who appears to be a public official or a professional. He is seated at a desk with a microphone in front of him, suggesting that he is either giving a speech or participating in a press conference. The man is wearing glasses and is looking down at some papers, indicating that he is reading or reviewing information. The background shows other people, likely fellow officials or journalists, seated in a row of chairs, which implies that this is a formal event or meeting. The style of the video is professional and formal, with a focus on the man and his actions. The video likely includes a news ticker at the bottom of the screen, providing additional information or context about the event or the man's speech."
+    ],
+    "video_ids": [
+        "H2sNHYx7uJA_2_24to160"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, a microphone, papers, and other people seated in chairs.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man in a suit and tie seated at a table with a microphone in front of him. He is handling papers, and other people are seated in chairs in the background. All core elements described are present in the video without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a man in a suit and tie, at a desk with a microphone in front of him, He is holding papers, which there people are seated in chairs in the background, The elements elements of in present and the video, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bcb96a72726dd9b8b9cb93424c2cd182be3795c8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6af69f7bac58f3958cb749e1037bc1955b0a7f17869d8d3cc9e3f4f1b1f2df58
+size 108855
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..86d1176e51ea57a2c1582857254d4a0f011d9376
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a journey on a snowy mountain road. The road is narrow and winding, with steep cliffs on one side and a river on the other. The car, a dark SUV, is seen driving along the road, its tires kicking up a spray of snow. The landscape is rugged and wild, with snow-covered mountains rising in the distance. The sky is overcast, casting a soft light over the scene. The road is icy, and the car's headlights cut through the gloom, illuminating the path ahead. The overall style of the video is realistic, capturing the harsh beauty of winter travel in a mountainous region."
+    ],
+    "video_ids": [
+        "eag0L7fTjbk_67_0to174"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A dark SUV driving on the road.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dark SUV driving on a snowy road, which matches the 'Object(s)' condition. The SUV is clearly visible and in motion, fulfilling the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a dark SUV driving on a snowy road, which align the descriptionObject(s)' condition. The vehicle is clearly visible and is motion, and the description description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..75a0419001d28f90c9d2ab811227c5ff3dcd5219
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b1bb23a96712105abe43177a3e3d9f80844337a3e24105f707535f44ae2ddb8
+size 209838
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..f59e413948b933ee5c8023f0d2200309c346d930
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a thrilling moment of a motocross rider in action. The rider, clad in a vibrant red and orange suit, is seen skillfully maneuvering a dirt bike over a sandy hill. The bike, with its black tires, stands out against the sandy terrain. The rider's helmet, matching the suit, adds to the overall dynamic scene. The video is shot from a low angle, emphasizing the height of the jump and the skill of the rider. The sandy hill, a common feature in motocross tracks, adds a sense of adventure and excitement to the scene. The rider's position on the bike and the angle of the jump suggest a high level of skill and control. The video is a dynamic and exciting portrayal of motocross, capturing the thrill and skill involved in the sport."
+    ],
+    "video_ids": [
+        "cIDMGHmYbfI_11_203to372"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A motocross rider in a vibrant red and orange suit, riding a dirt bike.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a motocross rider wearing a vibrant red and orange suit, riding a dirt bike across a sandy terrain. The rider's attire and the motorcycle are clearly visible and match the description provided. The background elements do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a motocross rider wearing a vibrant red and orange suit, riding a dirt bike. a sandy terrain. The rider's attire and the dirt are clearly visible, match the description provided. The environment,, not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..44f1a91a5b02d678928ef1258106f09d87cacf2b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:18cc6ceb203d4d90c1cfed393a2da6a254c5abc59461da8920c82207b2b992c8
+size 506100
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..ffb1236f8722d4a80b9cd2b1c3fce3acb48b692c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment featuring a woman named Brandi Glanville. She is seated in front of a backdrop with logos for KIIS FM and iHeart Radio. The woman is wearing a black top and has blonde hair. She is looking to the side with a slight smile on her face. The style of the video is a standard news interview, with a focus on the woman being interviewed. The backdrop suggests that the interview is taking place in a radio station or a studio associated with KIIS FM and iHeart Radio. The overall tone of the video is professional and informative."
+    ],
+    "video_ids": [
+        "L1QZyLo3TTU_14_71to203"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman named Brandi Glanville, a black top, blonde hair, and the KIIS FM and iHeart Radio logos.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair wearing a black top, identified as Brandi Glanville by the on-screen text. The background prominently features the KIIS FM and iHeart Radio logos, matching the specified conditions. Additional elements like the 'Clevver News' branding do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with blonde hair wearing a black top, which as Brandi Glanville. the text-screen text. The background includes features the logosIIS FM and iHeart Radio logos, which the description elements. The elements like the textKE'ver'' logo and not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..27c92af4250c50565c896fe7255f4145ead0c341
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0d401d4ccc6d3c06f14c79bdeedba197e53802c38bfbdb81fa6db48c2417a23a
+size 137025
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..7f6a86929b94ab1ef3ec49cf3b4b5289e666834f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a chocolate cake being decorated with candy. The cake is placed on a white plate, and the candies are scattered on top of the cake. The candies are orange and yellow, and they are placed on top of the chocolate frosting. The cake is then drizzled with chocolate sauce, which is poured over the top of the cake. The sauce is thick and glossy, and it drips down the sides of the cake. The cake is then placed on a table, and the camera zooms in on the cake, showing the details of the decoration. The video is a close-up shot, focusing on the cake and the decoration. The style of the video is a food video, showcasing the process of decorating a cake."
+    ],
+    "video_ids": [
+        "aVv_1XJVSLc_24_0to139"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chocolate cake, white plate, candies, chocolate sauce? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chocolate cake with chocolate sauce dripping down its sides, placed on a white plate. A hand is seen sprinkling colorful candies (yellow, orange, and brown) on top of the cake. All the specified objects are present and clearly visible in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a chocolate cake with a sauce dripping down, sides, placed on a white plate. There hand is seen addingling candies candies onlikely and orange, and red) onto top of the cake. The the elements elements\u2014 present and the visible in the video,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..29259e277b5c02e6aee12b251d2b2d8a1918cb9e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a5e22e2e0fc52b98bc5bac93c98c78cd01d352198acaf7104bdf6b5e62c650e
+size 209982
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..0b5ad6e1059026ff8d07cbf518b6251a67590f86
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a white and blue boat with a blue seat and white rails. The boat is on the water, and the seat is positioned in the center of the boat. The rails are white and have a blue design on them. The boat is moving through the water, and the seat is empty. The water is calm, and the boat is the only object in the frame. The boat is moving towards the right side of the frame. The seat is the main focus of the video, and it is shown in detail. The rails are also shown in detail, and they have a blue design on them. The boat is the only object in the frame, and it is moving through the water. The water is calm, and the boat is the only object in the frame. The boat is moving towards the right side of the frame. The seat is the main focus of the video, and it is shown in detail. The rails are also shown in detail, and they have a blue design on them. The boat is the only object in the frame, and it is moving through the water. The water is calm, and the boat is the only object in the frame. The boat is moving towards the right side of the frame. The seat is the main focus of the video, and it is shown in detail. The rails are also shown in detail, and they have a blue design on them. The boat is the only object in the frame, and it is moving through the"
+    ],
+    "video_ids": [
+        "YHBkNtfXwjE_32_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A white and blue boat with a blue seat and white rails.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a boat with a predominantly white interior and blue accents, including a blue seat and white rails. The visual elements match the description, with no contradictory features present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a white with a predominantly white body and blue accents, including a blue seat and blue rails. The design elements match the description provided indicating the additional details present.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..aa174586df398c3d952ff5641875792c59f57bc2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f261083f2c7aea8e1411fe35c6dbd238fb93d1c1754bb05c9fac5a9377f869d9
+size 205652
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..038c521beca9caae605c991d6a51304a8045f3e8
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a young man in a garage setting, with a casual and relaxed demeanor. He is wearing a black t-shirt and has short, light brown hair. The garage is filled with various tools and equipment, including a workbench, a metal cabinet, and a red chair. There are also some objects on the floor, such as a white ball and a red bucket. The lighting in the garage is dim, with a warm glow coming from the workbench area. The man appears to be in the middle of a project or task, as he is looking down at something in his hands. The overall style of the video is candid and informal, capturing a moment in the man's daily life."
+    ],
+    "video_ids": [
+        "Nnnww7q2d60_6_35to239"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young man, a black t-shirt, short, light brown hair, a workbench, a metal cabinet, a red chair, a white ball, and a red bucket.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a young man with short, light brown hair wearing a black t-shirt, standing in what appears to be a garage or workshop. In the background, there is a workbench, a metal cabinet (likely a tool chest), a red chair, a white ball, and a red bucket. All elements mentioned in the description are present and consistent with the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a young man wearing short, light brown hair wearing a black t-shirt. standing in a appears to be a workshop or workshop. There the background, there is a workbench with a metal cabinet,possibly a cabinet cabinet), a red chair, a white ball, and a red bucket. The these in in the description are present in match with the scene content of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..faec01325c24d9af50ed62c911fd1614dc45b930
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1391-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:66932cf0ef6897c44b223d0aa212c9e22c90651014095ee6e0b5891f64976cfe
+size 126116
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..158e5bdbeb33eaf9f1a46bcbeb8065c840eec964
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a sandy environment where a small insect, likely an ant, is actively engaged in digging or moving sand particles. The scene is dominated by a textured surface of fine, light brown sand grains, which vary slightly in color and size. The insect, positioned centrally in the frame, is seen using its legs to push sand particles aside, creating a small disturbance in the sand. As the insect moves, it occasionally lifts its antennae, possibly sensing its surroundings or communicating with other insects. The background remains consistent throughout the frames, emphasizing the detailed texture of the sand and the focused activity of the insect. The BBC logo in the top left corner suggests that this footage is part of a documentary or educational program. There are no significant changes in lighting or camera angles; the focus remains steady on the insect's actions, providing a clear and detailed observation of its behavior in this natural setting."
+    ],
+    "video_ids": [
+        "e88da73f84bd9f52551a9a5788ec3617b53fe6dc70b82c08a758ee06012f361d"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A small insect (likely an ant) and a textured surface of sand grains.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a small insect, consistent with an ant, interacting with a textured surface composed of sand grains. The ant is seen moving and digging into the sand, which aligns with the described condition. There are no conflicting elements that contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a small insect, likely with the ant, moving with a textured surface that of sand grains. The focus is moving moving through navigating through the sand, which fuls with the description conditions. The are no additional elements in contradict the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": Yes,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..69f36e736832cd7938ac56ba1be1485e8378b873
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:38c84423f6e16641feec96e470ecd0873a03c110959a900e534bac1228062aad
+size 201628
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..52b40ca9a6a9fc05603df31317145a846008b6ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a dynamic and colorful scene from a racing game. The main focus is a purple and yellow sports car with a large rear wing, driving down a highway. The car is in motion, with its tires leaving a trail of smoke on the road. The highway is set against a backdrop of a mountainous landscape, with a tunnel visible in the distance. The sky is clear and blue, suggesting a sunny day.  In the top right corner of the screen, there is a small window showing a young man's face. He appears to be a gamer, wearing headphones and a pink shirt. He is smiling and seems to be enjoying the game.  The video is a combination of in-game footage and a live stream of the gamer's reaction. The style of the video is fast-paced and action-packed, with a focus on the high-speed racing and the excitement of the gamer. The graphics are realistic and detailed, with a high level of detail in the car and the environment. The overall impression is of a thrilling and immersive gaming experience."
+    ],
+    "video_ids": [
+        "dR61LrLbrNc_6_235to434"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Purple and yellow sports car with a large rear wing, young man in the top right corner of the screen.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a purple and yellow sports car with a large rear wing, which matches the description. Additionally, a young man is visible in the top right corner of the screen, also matching the description. The presence of other elements, such as the race track and other cars, does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features a purple and yellow sports car with a large rear wing, which matches the description. Additionally, there young man wearing visible in the top right corner of the screen, also align the description. The presence of the elements like such as the tunnel track and tunnel vehicles, does not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0530b7e070c8061c26da25d5b781db9479397b67
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b6ad880ae9e7476ae6218a9d14369987332327faf501f58d3a14bd184807b4c
+size 241493
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..2f326031ca7be5a3027c696d5c0423846baf6747
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a blue and purple animated character is seen hugging a brown and pink animated character. The brown and pink character is holding a green plant. They are standing in a room with a pink and purple color scheme. In the background, there is a window with a blue and white pattern. The room also has a vase with purple flowers. The characters are standing on a pink floor. The video is in a cartoon style."
+    ],
+    "video_ids": [
+        "gdRE41y69rk_161_0to186"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two animated characters (blue and purple, brown and pink), a green plant, a vase with purple flowers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features two animated characters: one blue/purple (the adult figure) and one brown/pink (the child figure). A green plant is visible in the background, and there is a vase with purple flowers (or purple flower-like structures) on the right side of the scene. These elements align with the description, even though the vase is stylized and the flowers are not traditional.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows two animated characters, one blue andurple andleft character)) and one brown/pink (the child figure). They green plant is held in the hands, and there is a vase with purple flowers,though similar objects-like objects) in the floor side of the frame. The elements match with the description provided making though the characters is notized and not flowers are not clearly flowers}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3d6cdb1e4847c01c9ec2fa62de9e40fbfd581c47
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11f33c03ba257b2f4633f1d0817f6b0389f8b9e06bf89b1479b3948ddf052f85
+size 207222
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..63adc01261687f7c8c7b30d00b26013a6f49c4a6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two furry monsters are in a kitchen, preparing to bake cookies. The first monster, wearing a white chef's hat, is standing at a sink, washing his hands. The second monster, wearing a blue chef's hat, is holding a cookie sheet with several uncooked cookies on it. They are both standing in front of a counter with various kitchen utensils and ingredients. The kitchen is well-equipped with a sink, a stove, and a refrigerator. The scene is set in a brightly lit kitchen with a warm and inviting atmosphere."
+    ],
+    "video_ids": [
+        "EWwPFOQd2hw_22_71to215"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two furry monsters, one with a white chef's hat and the other with a blue chef's hat.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features two furry monsters, one pink and one blue, both wearing white chef's hats. The description specifies 'one with a white chef's hat and the other with a blue chef's hat,' which is slightly misleading since both are wearing white hats. However, the blue monster is clearly identifiable as blue, and the pink monster is pink, so the core description of two furry monsters with chef hats is fulfilled. The presence of the blue color on one monster satisfies the intent of the description, even if the hat color is inaccurately described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two furry monsters in one wearing and one blue, both wearing chef and's hats. The pink accurately thatone with a white chef's hat and the other with a blue chef's hat,' which is accurately inaccurate as both monsters wearing white hats. However, the core hat is wearing wearing as wearing, and the pink monster is wearing, fulfilling the core condition of two monsters monsters with chef's is met. The presence of additional hats monster on the of does the condition of the description.\"\n even if it color color is notately described.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..35d6673bf1e54a1a0a56c2133e07d68940cc80a1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e69a839614d3031ab45bba33a24742e8741f8f7789ea681c377e59cf8b40b3f
+size 217772
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..366f37fc67c5b68cbb5daf0ec148b84a4ff12d91
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a tall, stone church tower with a clock on its side. The clock is black and gold, and it's located halfway up the tower. The tower is topped with a battlement, and there's a bell above the clock. The sky is cloudy, and the sun is shining through the clouds, casting a warm glow on the tower. The tower is surrounded by trees, and there's a small building in the foreground. The video is in color and has a realistic style."
+    ],
+    "video_ids": [
+        "APZCfmgzoS0_25_0to196"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Church tower, clock, battlement, bell, trees, small building? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a church tower with a visible clock, battlement along the top, and surrounding trees. Although a bell is not clearly visible, its presence is plausible given the structure of a church tower. The small building is also implied by the visible church structure. All elements align with the description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a church tower with a clock clock, whichlements, the top, and a trees. There there bell is not explicitly visible, the presence is implied given the context of the church tower. The small building in also visible by the architectural part structure. The elements mentioned with the description provided contradict.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0d0687a2466c9666610c51b19b0348ab50f917a6
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28d6e67b88aa5d09a1eb7955cf9d374883a0f945b832c23763e1b3f92781483d
+size 163860
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..9b55deb61158c79971abc37a7dae6f91f3df2c69
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a group of five hikers on a rocky cliff, enjoying the view and the adventure. The hikers are dressed in casual hiking attire, with some wearing hats and bandanas, and they carry backpacks and water bottles. The cliff they are standing on is rugged and rocky, with a steep drop-off, and the sky is clear and blue. The hikers are spread out across the cliff, with some standing closer to the edge and others further back. They appear to be taking a break from their hike, possibly to take in the view or to rest. The overall style of the video is a mix of adventure and exploration, with a focus on the natural beauty of the landscape and the camaraderie of the group."
+    ],
+    "video_ids": [
+        "CfQc6LFftHc_39_0to108"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Five hikers? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows five individuals on a rocky overlook, dressed in hiking attire with backpacks, consistent with the description of five hikers. Their positions and actions (standing, sitting, taking photos) align with typical hiking behavior, and no elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts five h standing a rocky out, which in hiking attire, backpacks, which with the description of ' hikers. The positioning and the,standing and looking, and photos) align with the hiking behavior, and the additional contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a11db915a1f4a9a81b52bd4a0c6c0b809fc7b1ec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b8afdcec324c26226749908dc8ba4c98377c02d77cca47111b63c0ebb28c37d
+size 192494
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..27dd651e81e12c8b7181d3975e3435cd26c2bad5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are seated at a dining table in a restaurant. The man is holding a fork and is about to take a bite of his meal, which consists of a plate of fries and a side of salad. The woman is looking at the man with a slight smile on her face. The table is set with various items such as a bottle of ketchup, a glass of water, and a bowl of salad. The restaurant has large windows that offer a view of the street outside, where cars are parked. The overall style of the video is casual and intimate, capturing a simple yet meaningful moment between the two individuals."
+    ],
+    "video_ids": [
+        "FkON0734S50_22_0to185"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a woman, a dining table, a plate of fries, a side of salad, a bottle of ketchup, a glass of water, and a bowl of salad.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a woman seated at a dining table, eating. The man has a plate of fries and salad, and there is a bottle of ketchup, a glass of water, and a bowl of salad visible on the table. All specified objects are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man and a woman seated at a dining table. with. The table is a plate of fries and a, and there is a bottle of ketchup, a glass of water, and a bowl of salad on on the table. The the objects are present and correctly with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..75a09b33e0477dda8dd843da197f110bd8ef5b4b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6d17ded7fc07cca49b5e5656d79b6e38be8b266a4414af3d859e28c35150c82
+size 144862
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..9828d58ea3d919ae955cbad711710067fb50e7b5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a close-up of a single, illuminated light bulb with a visible filament. The bulb is positioned in the center of the frame, with a blurred background that consists of a series of colorful geometric shapes, possibly triangles, in various hues. The style of the video is artistic and abstract, with a focus on the interplay of light and color. The light bulb serves as the main subject, while the vibrant background provides a contrasting visual element. The overall effect is one of simplicity and elegance, with a sense of depth and dimension created by the interplay of light and shadow."
+    ],
+    "video_ids": [
+        "Pk6zchYgOlQ_8_0to173"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A single illuminated light bulb with a visible filament.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a single illuminated light bulb with a visible filament structure inside the glass bulb. The filament is clearly discernible as thin, yellowish rods extending from the base, which is consistent with the description. The background, though colorful and blurred, does not contradict the core object description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a single illuminated light bulb with a visible filament... bulb bulb. The filament is clearly visibleible, a, verticalish wires within from the central of which is characteristic with the description of The background is while colorful and abstract, does not detr the focus description being.}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b83a599fb2d0f96aaa0d61794f38f2a7cfa623ae
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1392-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:030b22978c99ed131e1310931d3d3e4555562dbce8820bb2a5154d7f9001f9ad
+size 77303
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..cbea53d708449d76e310cc213d5d34d8159db66d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a brown bear in a grassy field, surrounded by yellow flowers. The bear is seen walking through the field, its fur contrasting with the vibrant green of the grass and the bright yellow of the flowers. The bear's movements are slow and deliberate, suggesting a calm and peaceful environment. The field appears to be lush and well-maintained, with the flowers scattered throughout, adding a touch of color to the scene. The bear's presence in the field suggests that this might be a wildlife reserve or a protected area where bears are known to roam. The overall style of the video is naturalistic, capturing the bear in its natural habitat without any human intervention. The focus is on the bear and its surroundings, with no other objects or people visible in the frame. The video does not contain any text or additional elements, allowing the viewer to fully immerse themselves in the scene."
+    ],
+    "video_ids": [
+        "JqHWqZ26Ou0_76_275to415"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A brown bear? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a brown bear in a grassy field with yellow flowers. The bear is the central subject and its features, such as its brown fur, size, and behavior (eating grass), are consistent with the description of a brown bear. There are no elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features depicts a brown bear in a naturaly field with yellow flowers. The bear is the central object of matches features, such as its fur fur and ears, and shape (walking or), align consistent with the description of a brown bear. The are no elements in contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d7bcade9321c4c16689d9868c77afa3eb406d80e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fbb7023fbe53a5b74a169854c50a9c27726e51011024d8e5fefcff7b2c742e48
+size 284597
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..9fde1368d91318b9c18bbb33b6047d5b3c6d393d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is an aerial view of a house and its surroundings. The house has a red roof and is surrounded by trees. In the first frame, the house is shown from a distance, with the trees providing a natural backdrop. In the second frame, the camera zooms in on the house, revealing more details such as the red roof and the surrounding trees. In the third frame, the camera zooms out again, showing the house in relation to the surrounding landscape. The style of the video is realistic and it captures the beauty of the house and its surroundings."
+    ],
+    "video_ids": [
+        "-VtbzCn3Eq4_52_0to174"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: House, trees? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a house with a red roof and surrounding trees, which matches the 'Object(s)' condition. The aerial perspective captures both the house and the trees prominently, fulfilling the requirement without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a house with a red roof and white trees, which align the 'Object(s)' condition. The presence view and the the house and the trees,, and the requirement of any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..390ed99067d039706ab50cb915d884b2132eb092
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2dc23cb0e74565a095f8d6dc6ab0825e56bf81a90c300c1ccf0a0c4a68533f15
+size 234457
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..eb017a56858db466a5d6d8030ba92f76522265ff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a lively soccer practice session on a large, well-maintained grass field. The field is marked with white lines, indicating the boundaries and goals. Several soccer nets are set up across the field, and players are scattered around, engaged in various drills and exercises. Some players are seen running, while others are practicing their ball control and passing skills. The players are dressed in athletic attire, and the scene is filled with energy and movement. The field is surrounded by a neatly trimmed lawn, and the overall atmosphere is one of organized chaos, typical of a busy soccer practice session."
+    ],
+    "video_ids": [
+        "cPYduVWocSE_3_0to102"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Players, soccer nets? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows multiple players on a soccer field, actively engaged in training or practice, and several soccer nets are visible positioned across the field. These elements align with the 'Object(s)' condition of players and soccer nets.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts shows players players on a soccer field, engaging engaged in a or a. and there soccer nets are visible, around the field. The elements align with the 'Players(s)' condition, the and soccer nets,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7dd1cc4e600613a1036ed11536769b3a55efa285
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0f2858d437d3aad5cc0271ecf9aaf6f830b40abd4fb30f044c18b3001168fb0
+size 248649
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..f9c69eb94f2c43bb3236c02e7cbea4103f23df86
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man wearing a black beanie and sunglasses, standing in a grassy area with a road and buildings in the background. The man is smiling and appears to be enjoying the outdoors. The sky is cloudy, suggesting it might be a cool or overcast day. The man is dressed in casual attire, including a black jacket, and seems to be in a relaxed and happy mood. The overall style of the video is casual and candid, capturing a moment of the man's life in an everyday setting."
+    ],
+    "video_ids": [
+        "uKsUpND6DfQ_11_730to854"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: man wearing a black beanie and sunglasses, dressed in a black jacket? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video consistently shows a man wearing a black beanie and sunglasses, dressed in a black jacket. The core description matches the visual content without contradiction, even though other elements like the background and lighting are present.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man wearing a black beanie, sunglasses, dressed in a black jacket. There background elements is the visual elements of any or and though the elements like the background ( the may not.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f6cb3f8353af9a4b069d4aac655424021c8b7368
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57caf01cdb171dacd88f1837fd1add16e6641ea6d6253b4a3b67d1e95ccdad47
+size 157042
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..8a5aaf8443c45bf8bd27645270d4d03f7381283a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a vibrant underwater scene, featuring a school of yellow fish swimming through a coral reef. The fish, with their bright yellow bodies and black stripes, are the main focus of the video. They are seen swimming in different directions, some closer to the camera and others further away, creating a sense of depth and movement. The coral reef, with its various shades of purple and orange, provides a beautiful and colorful backdrop for the fish. The video is shot from a low angle, looking up at the fish and the coral, which adds to the sense of scale and grandeur of the scene. The overall style of the video is realistic, capturing the beauty and diversity of marine life in its natural habitat."
+    ],
+    "video_ids": [
+        "UJM5TAe8GqU_23_0to134"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A school of yellow fish with black stripes.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a school of yellow fish with distinct black stripes, which matches the description. These fish are the main focus and are clearly visible swimming among coral and other marine life. The presence of other fish and coral does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows features a school of yellow fish with black black stripes, which align the description provided The fish are swimming central focus of are swimming visible against in coral. other underwater elements. The presence of additional elements and marine does not contradict the core description but}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1c67052a28d785bba571dd646f2ff013ba558a16
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9d7d3aefcf02e705fef8e1075ed13936ee101368dbfca78136743baf84f39736
+size 208037
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..e6253fa45c38a77153a1402d614c013284ffdd9a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a car's dashboard and center console. The dashboard features a digital display showing the car's battery level, fuel level, and other information. The center console has a touch screen display, which is being interacted with by a person's hand. The car's air conditioning vents are visible, and the car's interior is well-lit, suggesting it is daytime. The style of the video is a straightforward, real-life depiction of a car's interior, focusing on the technology and controls available to the driver."
+    ],
+    "video_ids": [
+        "Itb0C8_2-PA_9_816to962"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Digital display on the dashboard, touch screen display on the center console, a person's hand interacting with the touch screen, car's air conditioning vents.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a digital display on the dashboard (instrument cluster) and a touch screen display on the center console. A person's hand is visible interacting with the touch screen, and the car's air conditioning vents are prominently displayed below the screen. All elements described in the condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a digital display on the dashboard,top cluster), and a touch screen display on the center console. A person's hand is interacting interacting with the touch screen, and the car's air conditioning vents are also displayed and the touch. The elements described in the question are present and accurately depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1cf1fcaa57ee74b995237426bbd50cfdc64eb9ec
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:163c302a186be61efe96eca1d1b1d9a2e84f037777ae3c1dce12db9eedd4e5cf
+size 202849
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..e665675533571822bf698dc820e88e304124ea58
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a person is seen interacting with a hedgehog in a grassy area. The hedgehog is brown and has a spiky coat. The person is gently petting the hedgehog, which is lying on the ground. The hedgehog appears to be comfortable with the person's touch. The background of the video shows a grassy area with a white pole. The overall style of the video is calm and gentle, capturing a moment of interaction between a human and a wild animal."
+    ],
+    "video_ids": [
+        "BY_L-NZUcPM_17_0to181"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A person, a brown hedgehog with a spiky coat.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person's hand gently touching a brown hedgehog with a spiky coat, which matches the description. The hedgehog is on grass, and while there are other elements in the background (like a white pipe), they do not contradict the core description of the person and the hedgehog.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person's hand interacting pet a brown hedgehog with a spiky coat. which align the description of The hedgehog is in a, and the there are other elements like the background,green green white pole and they do not conflict the core description of the person and the hedgehog.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..90908c622f0eff325ba719f8e4a731c5a638346c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:16b8499f73db66378c0da9911ee7e4990f24cc95c3b00be373f439c9c02cf022
+size 303165
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..f4ea4d9ca2c21b03c580b9745d94fc3527a50978
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a young couple is seen enjoying a coffee date in a cozy caf\u00e9. They are seated at a small white table, each holding a cup of coffee. The man is wearing a purple and white sweater, while the woman is dressed in a pink and white sweater. The caf\u00e9 is decorated with potted plants and framed pictures on the wall, creating a warm and inviting atmosphere. In the background, other patrons can be seen enjoying their meals and drinks. The video captures the intimate moment between the couple, as they engage in conversation and enjoy each other's company."
+    ],
+    "video_ids": [
+        "5wm0n90hNzg_6_0to122"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A young couple, a small white table, cups of coffee, a man in a purple and white sweater, a woman in a pink and white sweater, other patrons.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a young couple (a man in a purple and white sweater and a woman in a pink and white sweater) seated at a small white table with cups of coffee. Other patrons are visible in the background, and the setting appears to be a caf\u00e9 or similar environment. All core elements from the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video depicts ful a young couple sittinga man in a purple and white sweater and a woman in a pink sweater white sweater) sitting at a small white table with cups of coffee. The patrons are not in the background, which the setting appears to be a cozy with a environment, The elements elements of the description are present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3da96cff81e8b41d0718f058827e427ca9ff07e5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1393-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ef00292cc86daec5f8efdef45db74cedf6ebe2c3ac9245c314933233449c65d
+size 197578
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..243949d1d9f60ce7e9b1b8b777475562e3862415
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a woman in a blue shirt is engaged in a conversation with two men in a kitchen setting. The woman is gesturing with her hands as she speaks, indicating an active discussion. The man on the left is wearing a red apron and a black hat, suggesting he might be a chef or a cook. The man on the right is wearing a blue jacket and a red bandana around his head, possibly indicating a casual or informal setting. The kitchen is well-equipped with various appliances and utensils, including a refrigerator, an oven, and a sink. The overall style of the video suggests a casual, friendly interaction between the three individuals, possibly discussing a recipe or a cooking technique."
+    ],
+    "video_ids": [
+        "WYipNMklDkA_81_17to200"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in a blue shirt, a man in a red apron and black hat, and a man in a blue jacket with a red bandana.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three individuals matching the described attire: a woman in a blue shirt, a man in a red chef's jacket (which can be interpreted as a red apron in context) and a black hat, and a man in a blue jacket with a red bandana. All core elements are present and accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a individuals in the description objects: a woman in a blue shirt, a man in a red apr's apr andwhich can be considered as a red apron) the) and a black hat, and a man in a blue jacket with a red bandana. The elements elements of present and accurately represented.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..efd9606b755332d40b233c1bc45acdc152793a76
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:320fb34ad7eca363432a9345701f42f9135f10bf123b6adb664612d7bfaabe01
+size 184042
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..d29050d64a2005c0415c4cdc100f75f0dcd49e55
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a group of people enjoying a boat ride on a calm river. The boat, painted in vibrant shades of blue and red, is filled with passengers who are seated comfortably. The man in the foreground, donned in a yellow shirt, is holding a camera, seemingly capturing the picturesque surroundings. The boat is moving towards a bridge in the distance, which adds a sense of depth and perspective to the scene. The overall atmosphere of the video is serene and leisurely, with the passengers appearing relaxed and content."
+    ],
+    "video_ids": [
+        "KJAuYWw0k24_153_0to112"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A boat, passengers, a man in a yellow shirt holding a camera, a bridge? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a boat with passengers, a man in a yellow shirt holding a camera (likely a selfie stick), and a bridge in the background. All core elements described are present in the video without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a boat with passengers, a man in a yellow shirt holding a camera,which a smartphone stick or and a bridge in the background. The the elements of in present, the video, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d80a5ae4012fedae361a3808a68f8929c0dbe38e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1005a2fa74c6fc1857f1cfc764808e9873f04cf0a26ca87fea8516096dbb64b9
+size 218298
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..9914d30f2803df91fbd2621365770907a22c960a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a group of soccer players walking across a field. They are dressed in blue jerseys with the word \"beko\" printed on them. The players are smiling and appear to be in good spirits. The field is green and well-maintained, with a clear sky above. In the background, there are stadium seats, suggesting that this is a professional soccer stadium. The style of the video is candid and informal, capturing a moment of camaraderie among the players."
+    ],
+    "video_ids": [
+        "kQeiBZXBijE_17_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A group of soccer players dressed in blue jerseys with 'beko' printed on them.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group of soccer players wearing blue jerseys with 'beko' printed on them, which matches the described condition. The players are on a field, and while there are additional elements like stadium seating and other players in the background, these do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a group of soccer players wearing blue jerseys with 'beko' printed on them. which matches the description '. The players are walking a soccer, and the there are additional elements like the seating and a players in the background, these do not contradict the core description of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7a86fa9d95d85e08ec2b27e1c9d2797e2bbd0a1b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e9327d428dda1fa3b3995aec42d9b4e0fd72a9ca783132adc36f850c4d8f4378
+size 273770
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..1e0ef4a7c91084315dc1355720651ed75ecf4ad4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a scenic drive along a winding coastal road. The road, which is the main focus of the video, is surrounded by lush greenery and rugged cliffs. The ocean, visible in the background, adds a sense of tranquility to the scene. The video is shot from a high angle, providing a bird's eye view of the road and the surrounding landscape. The colors in the video are vibrant, with the green of the trees and the blue of the ocean creating a striking contrast. The overall style of the video is serene and picturesque, capturing the beauty of nature and the joy of a road trip."
+    ],
+    "video_ids": [
+        "rOTvRjAxCkw_14_0to137"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Winding coastal road, lush greenery, rugged cliffs, ocean? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a winding coastal road running alongside the ocean, with lush greenery covering the surrounding cliffs and hills. The ocean is visible with waves crashing against rocky shores, matching the described elements. Additional text overlays do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a winding coastal road that parallel lush ocean, with lush greenery and the cliffs landscape. vegetation. The rugged is visible in its crashing against the cliffs, and the description elements. The elementsural or not contradict the core description and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..94dcb50fdcab308003211a90a0db0e3764182964
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30f347d640c22e41ba97b05b3b8fc7649d2b3c421652a84e6b23134ffa7bed55
+size 457188
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..412955c68f1c7f5c8fa60b043f894b21a30797fa
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a vibrant aquarium teeming with life. The tank is filled with a variety of colorful fish, including red tetras, blue and yellow guppies, and other small, brightly colored species. These fish swim gracefully among lush green aquatic plants that cover the bottom and sides of the tank. A prominent piece of driftwood serves as a central focal point, adding a naturalistic element to the scene. The water is clear, allowing for an unobstructed view of the fish and plants. Bubbles rise from the bottom, indicating the presence of an air pump, which helps oxygenate the water. The overall environment is serene and lively, capturing the beauty and tranquility of a well-maintained aquarium ecosystem."
+    ],
+    "video_ids": [
+        "91d080f7750711f952142d3caa061b8e04c754e7406e149855b42c0dee3e4689"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Red tetras, blue and yellow guppies, other small, brightly colored fish, aquatic plants, driftwood, bubbles.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by showing a variety of fish including red tetras (small, red fish), blue and yellow guppies (yellow fish with blue tails), and other small, brightly colored fish. It also features aquatic plants, driftwood, and bubbles rising from the filter, all of which are clearly visible and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting a variety of fish, red tetras,visible, brightly fish), blue and yellow guppies (small and with blue spots), and other small, brightly colored fish. The also features aquatic plants and driftwood, and bubbles, from the water, which of which are present visible in match with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7cc2824f475746aaade279db4a39dfe0108b0e8d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:043d4b376df10cc7b6f58e8af14c109aaec45ccd9b4c0740d62c88f56bdc38cf
+size 213760
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..bb413410923666fea994d0dcea1cb788687af3d0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a large, intricate circular design on the side of a building. The design is black and white, featuring a detailed pattern with a central figure that resembles a face or mask. The building itself is white with a curved roof, and the design is prominently displayed on the side of the structure. The style of the video is a straightforward, real-world shot with no additional effects or filters applied. The focus is on the design and the building, with no other significant elements or actions taking place. The video captures the design in its entirety, allowing viewers to appreciate the complexity and detail of the artwork."
+    ],
+    "video_ids": [
+        "9mKDxrKQxtE_8_84to229"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A large, intricate circular design, a central figure resembling a face or mask? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a large, intricate circular mosaic design on the facade of a building. The central figure within this circle clearly resembles a face or mask, composed of detailed patterns and shapes. This matches the described 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a large, intricate circular design design on the facade of a building. At central figure within the design is resembles a face or mask, with of detailed patterns and lines that The matches the description 'Object(s)' condition perfectly any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..f08d62010d4136c87b39e6973091fd4b7ee98057
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03ddf00fc4d84b09cfc0ec32b026f2141ee8dc857b0b47d41b7e9b5c7bea7957
+size 99292
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..95d83074180a5b1ccdbb7cb9562b5b70db30b31b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a sports talk show featuring three men in a studio setting. The host, wearing a purple shirt, is seated at a desk with a microphone, engaging in conversation with two other men. The first guest, wearing a suit, is seated to the host's right, while the second guest, wearing a black shirt, is seated to the host's left. The studio has a large screen in the background displaying the Super Bowl logo and the text \"LIVE ATLANTA\". The show is titled \"The Stephen A. Smith Show\" and the host's Twitter handle is displayed at the bottom of the screen. The overall style of the video is professional and polished, typical of a sports talk show."
+    ],
+    "video_ids": [
+        "5AccDDowo6k_10_620to759"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Host (wearing a purple shirt), First Guest (wearing a suit), Second Guest (wearing a black shirt)? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows three individuals at a sports talk show set. The host on the right is wearing a purple shirt, the first guest in the middle is wearing a suit, and the second guest on the left is wearing a black shirt. All three individuals match the described attire, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows two individuals seated a table commentary show set. The person is the right is wearing a purple shirt, and first guest on the middle is wearing a suit, and the second guest on the left is wearing a black shirt. The the individuals are the description attire, fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..1bc0d58b3a75f8e03e4515e4e45c3181d8afcfb0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e307dadca0dcc61d5163c432a56736f9364336182546a00a6f32672ffebbc7e3
+size 168657
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..6824d314a8a1810faf640f6b61f965a4fbfbe57f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a promotional advertisement for the new Mercedes-Benz truck. The style of the video is sleek and modern, with a focus on the truck's design and features. The truck is shown in a large, open space with a red and white background, which adds a dramatic effect to the presentation. The man in the video is standing confidently in front of the truck, wearing a gray sweater and a blue tie. He is speaking to the camera, likely discussing the truck's features and benefits. The overall tone of the video is professional and polished, with a clear emphasis on the truck's design and performance."
+    ],
+    "video_ids": [
+        "DWBYbPNUojU_0_0to112"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Mercedes-Benz truck, man in gray sweater and blue tie? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a Mercedes-Benz truck (specifically, a G-Class) and a man wearing a gray sweater and blue tie. The man is positioned in the foreground, and the truck is prominently displayed behind him, matching the described objects. Additional elements in the background (e.g., people, signage) do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a Mercedes-Benz truck inpartally a a white-Class vehicle in a man wearing a gray sweater and a tie. The truck is standing in front foreground, and the truck is in displayed in him, fulfilling the description elements. The elements, the background,such.g., the and other) do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..59a6d0bd59affc5fe37ab3145d68fb777b1ed2c5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1394-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:915e8f7e814e0eb52386529c8b13dc51e07f20cdfbfedae4dde098fc386e3e15
+size 106248
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..f212903685424019a8ff59aa994be056459f324a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a suit standing inside an airplane cabin. The man is smiling and appears to be speaking. The airplane cabin is well-lit, with natural light coming through the windows. The windows are round and provide a view of the sky outside. The man is standing in front of a row of seats, which are empty. The style of the video is candid and informal, capturing a moment in the man's journey."
+    ],
+    "video_ids": [
+        "OdzQ0Tb5DqY_3_0to121"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit, a row of empty seats? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man dressed in a suit standing inside what appears to be a private jet. Behind him, there is a row of empty, white leather seats, which matches the described elements. The setting and objects are consistent with the 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man in in a suit standing next an appears to be an commercial airplane cabin There him, there is a row of empty seats luxurious seats seats, which align the description '. The setting and the in consistent with the givenObject(s)' condition provided any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..bda6e7f43d87612c74fabdf5f519219bcaae988d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2da92dba2bdaaa5cc320d36782be9bbf0fd00949a8a998f0bc33969c542d3109
+size 143404
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..82fa037e8f39236c0a45a3c125feb73e780490b7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is carving a steak on a wooden cutting board at a dining table. He is using a knife and fork to slice the meat. Two other men are seated at the table, watching the process. The table is set with plates, forks, and knives. The room has a yellow wall and a window with blinds. The men are wearing casual clothing, and the overall atmosphere is relaxed and friendly."
+    ],
+    "video_ids": [
+        "Zn31bxSw3fI_37_677to806"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man carving a steak, two seated men watching, a wooden cutting board, plates, forks, and knives.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man actively carving a steak on a wooden cutting board, with two seated men watching him. Plates, forks, and knives are visible on the table, matching the described elements. The scene is consistent with the 'Object(s)' condition without any clear contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a man carving carving a steak on a wooden cutting board, with another other men watching.. The, forks, and knives are also on the table, which the description elements. The scene is set with the givenObject(s)' condition provided any contradictions contradictions.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e3778a0af471d74d454c228d94412d755248af67
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:12cf41f1cc7aa1a046ec2fd5a9c9acff4478c4d0960579442fc9ef241cc5f8c9
+size 209832
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..a6e1c856f4be7dad287b38777b77b8b4527ad739
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a stylized, artistic representation of a beer mug in front of a desert cityscape. The beer mug, filled with a frothy golden liquid, is the central focus of the image. The cityscape, depicted in a warm, earthy color palette, features ancient-looking buildings and palm trees, suggesting a location in a desert region. The overall style of the video is reminiscent of a travel advertisement, with the beer mug symbolizing a leisurely, relaxing experience in the depicted location. The image is likely meant to evoke a sense of adventure and exploration, combined with the enjoyment of a refreshing beverage."
+    ],
+    "video_ids": [
+        "cJ0dsYPOv78_24_0to166"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Beer mug filled with frothy golden liquid.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a beer mug filled with a frothy golden liquid, which matches the description. The mug is centrally positioned and clearly visible throughout the video, with bubbles rising in the liquid, indicating it is beer. The background elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a beer mug filled with fro frothy golden liquid, which matches the description. The liquid is placed placed, the displays, the frames. with the and to the liquid, indicating it is indeed. The background,, not detr the main description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..fdb2470ec179cf38c637af853350823a4a8bc30b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:32c65f0527fd0857ef8612073136a7c266928c621589ef999c21e15b12e71443
+size 170325
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..57e58ebd56969c74e822617c8cf0336fda7da8d5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a vibrant scene at an amusement park. A yellow roller coaster car, filled with excited passengers, is seen in motion, ascending a red track. The roller coaster car is the main focus of the video, with its bright yellow color contrasting against the red track. The passengers are dressed in casual attire, adding to the lively atmosphere of the amusement park. The roller coaster car is moving from the left to the right side of the frame, indicating the direction of the ride. The background of the video is filled with lush green trees and colorful umbrellas, creating a festive and joyful ambiance. The video is a dynamic representation of a day at the amusement park, filled with fun and excitement."
+    ],
+    "video_ids": [
+        "TksFucf7sUI_11_68to190"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Yellow roller coaster car with excited passengers.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a yellow roller coaster car with passengers riding it, which matches the description. The passengers appear to be excited, as evidenced by their body language and expressions, and the car is clearly yellow. Additional elements like the surrounding park and structures do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a yellow roller coaster car with passengers, on. which matches the description of The car appear to be excited, as indicated by their posture language and the. which the car is on on. The elements like the colorful environment and the do not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..058e9c1b4a28a2bc8b2c73fef5559ef81fa1b0b7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1788e497236e7d5b8879ddbbbc3117ace07ac10889745bbb88c03b2160629932
+size 260011
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..4e8e2ccef455101849fcd1c2bd6c5959ca5e8ecb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a news segment from MSNBC, featuring a man standing outdoors in front of a building. The man is wearing a checkered shirt and has a microphone clipped to his shirt, indicating that he is a reporter or news anchor. The background shows a clear sky and a building with a green awning. The text overlay on the video reads \"MUELLER REPORT IMPACT\" and \"KEY SWING STATE VOTERS WEIGH IN ON MUELLER FINDINGS\", suggesting that the segment is discussing the impact of the Mueller report on voters in key swing states. The style of the video is typical of news broadcasts, with a focus on delivering information in a clear and concise manner."
+    ],
+    "video_ids": [
+        "HQFhF42wYeM_2_69to194"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man wearing a checkered shirt with a microphone clipped to his shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a checkered shirt with a microphone clipped to his shirt, which matches the description. The background and additional elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a checkered shirt, a microphone clipped to his shirt. which matches the description provided The background appears text text, not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2fd2b82c26686a309adf6f3b504a954c67cba379
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6be564b04c3997788a3c77c9c06996285afb9dc26975e56e7ae15f030de5ab9
+size 225925
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..e453d2a5320e4e84ceb6504641fa19dbb89496bb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a woman in a bridal shop, standing in front of a mirror. She is wearing a white dress with a polka dot pattern and has her hair styled in an updo. The shop is filled with various wedding dresses on display, and there are pictures hanging on the wall. The woman is smiling and appears to be happy. The style of the video is a casual, candid shot, capturing a moment in the woman's life as she prepares for her wedding."
+    ],
+    "video_ids": [
+        "4dqE01EUyvA_11_0to200"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman, a white dress with a polka dot pattern, an updo hairstyle, various wedding dresses, and pictures hanging on the wall.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman wearing a white dress with a polka dot pattern, and she has an updo hairstyle. The setting is a bridal shop with various wedding dresses displayed on mannequins and hanging on racks, and framed pictures are visible on the walls. All core elements from the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a woman wearing a white dress with a polka dot pattern, which the is an updo hairstyle. The background appears a room shop with various wedding dresses on on racksnequins and racks on racks. and there pictures are visible on the wall. The elements elements of the description are present in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9c1693b260bdb5d6a7cc95eff4f5c83b826ab75d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90282e9a241e117621231afecfc028f3f94950ce8de3fdc00765505f9c76813f
+size 177613
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..143fc5dfb28f126f7c2dc8ec8f58dc2e2626f631
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a medical professional is seen taking a patient's blood pressure. The patient's arm is placed on a table, and the professional uses a blood pressure cuff to measure the pressure. The patient's hand is held by the professional, providing comfort and support. The blood pressure monitor displays the reading, which is 189. The scene takes place in a medical setting, with a white table and a window in the background. The professional is dressed in a blue shirt, and the patient is wearing a white shirt. The overall style of the video is clinical and professional, focusing on the medical procedure being performed."
+    ],
+    "video_ids": [
+        "L7z3VurExTM_12_0to105"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A medical professional, a patient, a blood pressure cuff, a blood pressure monitor, a white table, and a window.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully depicts a medical professional (hands in white coat), a patient (arm with blood pressure cuff), a blood pressure cuff, a blood pressure monitor displaying readings, a white table, and a window with blinds in the background. All core elements described are present and consistent with the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful a medical professional (w of a gloves) a patient (arm with blood pressure cuff), a blood pressure cuff, a blood pressure monitor, a, a white table, and a window in blinds in the background. The elements elements are in present and accurately with the scene.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..41b2021eab13f54f1dcdd5bc4d926eb0f607f049
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c080b10a15f738bd40fb3b0d0b3c8cbd21444ab4848442652f887d06d6c5a4b
+size 118622
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..05197242a9ce18fe0b6c70bfc967bbdc801cd9b9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a baby's journey from being asleep to waking up and smiling. The baby is seated in a car seat, which is gray and black. The car seat has a warning label on it, indicating safety precautions. The baby is dressed in a blue and white striped onesie, and a white blanket is wrapped around them. The baby's eyes are closed in the first frame, indicating that they are asleep. In the second frame, the baby's eyes are open, and they are looking directly at the camera, showing curiosity and alertness. In the third frame, the baby is smiling, indicating that they are happy and content. The video captures the baby's transition from a state of rest to a state of alertness and happiness."
+    ],
+    "video_ids": [
+        "DaJlYW6xY_g_6_0to188"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A baby in a blue and white striped onesie, a white blanket, a car seat, a safety warning label.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a baby wearing a blue and white striped onesie, wrapped in a white blanket, seated in a car seat. A yellow safety warning label is visible on the car seat. All core elements from the description are present without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a baby wearing a blue and white striped onesie, sitting in a white blanket, seated in a car seat. The safety safety warning label is visible on the car seat. The elements elements of the description are present, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..361340d5463083b6fc29963d4b3f6fdee466fcc9
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1395-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d9e39897ac2153fae6e4e8c0ebc027711587f1fc93970ef3edb93db7825dd3e1
+size 251155
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..5a4394a9aa291e9f51a6fc84bceab2d21eaa917e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen in a living room setting, standing in front of a window with orange curtains. He is dressed in a black and white striped robe, and a white t-shirt underneath. His right hand is raised, pointing towards the camera. The man appears to be in the middle of a conversation or making a statement. The room is well-lit, with a lamp visible in the background. The overall style of the video suggests a casual, everyday scene."
+    ],
+    "video_ids": [
+        "uef3eyg8vS0_1_0to142"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black and white striped robe and white t-shirt, standing in front of a window.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The man in the video is wearing a black and white striped robe over a white t-shirt, which matches the description. He is standing in front of a window with red curtains, and while there are additional elements like wall lamps and framed art, they do not contradict the core description. The video largely fulfills the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video in the video is wearing a black and white striped robe over a white t-shirt, and matches the description. He is standing in front of a window with curtains curtains, which there the is additional elements like a lamps and a pictures, these do not contradict the core description. The setting largely fulfills the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3ab8c427f6dd90e34bd01c21b7e9e07a6d07c06d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:15bf6545f824833fea05a2672bbc24e7d6db59f149a49ce430d3cf3c4b8fd689
+size 359899
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..94c8d7bdc61e3a652271f0db3cbcb69d4cb9b6bc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a black jacket with a blue patch on the sleeve, pointing to the right. He is standing in front of a white wall with horizontal lines. The man appears to be in a sports setting, possibly a coach or a manager, as he is dressed in a sports jacket with a logo on the left chest. The style of the video is a simple, straightforward shot of the man in action, with no additional elements or distractions. The focus is solely on the man and his gesture."
+    ],
+    "video_ids": [
+        "wNFN9M_lvAs_12_82to290"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black jacket with a blue patch on the sleeve, a white wall with horizontal lines, and a logo on the left chest.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black jacket with a blue patch on the sleeve, standing in front of a white wall with horizontal lines. The man also has a logo on the left chest of his jacket, matching the described elements. There are no conflicting elements in the scene that contradict the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a black jacket with a blue patch on the sleeve, standing in front of a white wall with horizontal lines. The logo is has a logo on the left chest of his jacket, which the description elements. The are no additional elements in the video.\"\n contradict the given.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b58b7959b6f841beae907fc4332e2d3a691be43e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb4d33a772109a29e71ba5b82e82f76132bd147e59b22898e26b4aecce20a1b9
+size 121986
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..ee35875580adca953c3da735319f2594d7e1e067
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a chef is preparing a dish in a kitchen. The chef is wearing a black apron and white gloves, indicating a professional setting. The chef is using a spoon to mix ingredients in a glass bowl. The bowl is placed on a green cutting board, which is on a black countertop. The cutting board also has a lemon and a lime on it. In the background, there is a silver pot and a pink cloth. The chef's actions suggest that they are in the process of making a dish that requires mixing and possibly adding citrus flavors. The overall style of the video is a close-up shot of the chef's hands and the ingredients, focusing on the cooking process."
+    ],
+    "video_ids": [
+        "SYH8TRHBATQ_11_27to230"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Chef, spoon, glass bowl, lemon, lime, silver pot, pink cloth? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a chef wearing gloves, using a spoon to add a green paste from a small glass bowl into a larger glass bowl containing meat. Two lemon halves are visible on a green cutting board. A silver spice container with multiple compartments is present, and a pink cloth is in the foreground. All listed objects are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a person wearing gloves, using a spoon to mix ingredients liquid substance into a silver silver bowl into a larger glass bowl containing a and There le and and visible on the cutting cutting board, The pink pot container and a compartments is also, but a pink cloth is also the background. The these objects are present and correctly with the scene.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2f99753d408cd7f74ae14186c418f18238ed2214
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:32e2a2462112440ea7e928f7c8f68f4ffb55e19834fb78c5f3d36c39f07f14ad
+size 139096
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..428c69f2b9c9d7ff89fd14e7ac71ca09eadd67cb
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with long brown hair, wearing a yellow top. She is seated in front of a red background with purple accents. The woman appears to be in a studio setting, possibly for an interview or a talk show. She is looking to her left with a slight smile on her face, suggesting a relaxed and friendly demeanor. The lighting in the studio is bright, highlighting the woman's features and the vibrant colors of her top and the background. The overall style of the video is professional and polished, with a focus on the woman's expression and the colorful studio environment."
+    ],
+    "video_ids": [
+        "0AUac4z2HTU_11_44to207"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with long brown hair wearing a yellow top, seated in front of the background.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with long brown hair wearing a yellow top, seated in front of a background with orange and purple hues, which matches the description. The core elements are present and consistent throughout the frames.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with long brown hair wearing a yellow top, seated in front of a background with horizontal and blue stripes. which matches the description provided The presence elements of present and consistent with the frames.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..e7470ffe7ae5457465adb31c38836d0930dc01d3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85f6c9db8bc006868e5694af3cf57b0c115d50c24752cd1003c1444d87698b34
+size 113161
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..bb8ae9864f42e8c93c9520568be2aacce8192088
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the vibrant beauty of a field of yellow flowers in full bloom. The flowers, with their bright yellow petals and green stems, are the main focus of the video. They are arranged in a way that creates a sense of depth and perspective, with some flowers appearing closer to the camera and others further away. The flowers are set against a backdrop of lush green grass, which adds a contrasting color to the scene. The video is shot in a way that allows the viewer to appreciate the details of the flowers, from the texture of the petals to the way they sway in the breeze. The overall style of the video is naturalistic, with a focus on capturing the beauty of the flowers in their natural environment."
+    ],
+    "video_ids": [
+        "arkBoag1z6Y_1_0to110"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Yellow flowers, green grass? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features yellow dandelion flowers and green grass, which aligns with the specified 'Object(s)' condition. The flowers are clearly visible, and the background consists of green grass, fulfilling the core description without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video primarily features yellow flowersaisyion-like, green grass, which aligns with the ' 'Object(s)' condition. The presence are in visible in and the green includes of green grass, fulfilling the description description without any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a63cfcfd1b7174ebeab0a0ca8aba7b532bdf833a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:65af214b4ee106577637bce28a1aae8afa9cfb6c14015e6a15761dd91d020815
+size 166502
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..7d468a58322095121bbf3eb8dd7be1fbccefbc9b
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two young men are seen in a room with a vintage aesthetic. The first man, wearing a blue shirt and glasses, is standing in front of a whiteboard with a map on it. He is smiling and appears to be explaining something to the second man, who is wearing a tie-dye shirt and a baseball cap. The second man is looking at the map on the whiteboard, seemingly engaged in the conversation. The room is filled with various objects, including a green desk, a red cabinet, and a potted plant. The overall style of the video is casual and friendly, capturing a moment of interaction between the two men in a comfortable and inviting setting."
+    ],
+    "video_ids": [
+        "EbzwyYCcijA_22_0to109"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two young men, a whiteboard with a map, a green desk, a red cabinet, and a potted plant.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows two young men interacting in a room that contains a whiteboard with a map (partially visible), a green desk with various items on it, a red cabinet with glass doors, and a potted plant on a shelf. All specified objects are present and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts ful two young men in in a room. includes a whiteboard with a map,partially visible), a green desk, a items on it, a red cabinet, a doors, and a potted plant in the shelf. The the objects are present and contribute with the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..90cf537ef941aa6e7b199bf41906114a062213b7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9720bb2f3e52f9b179d88521da4005897b2f47f85d9a697cf50c5e116b2cf20c
+size 176117
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..108d84942fa1b7f151dbd2956f6afde1e3bb8221
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man in a black and red shirt standing next to a white Ram truck. He is pointing at the truck, possibly explaining something about it. The truck is parked on a grassy area near a body of water, which could be a lake or a river. The man is bald and wearing sunglasses. The truck is a four-door model with a chrome grille and silver rims. The man appears to be in a sales or demonstration role, as he is gesturing towards the truck with confidence. The setting is serene and natural, with the water and the greenery providing a peaceful backdrop for the scene."
+    ],
+    "video_ids": [
+        "GhEphnYEVe4_12_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a black and red shirt and a white Ram truck.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing a black and red shirt standing next to a white Ram truck. The man is positioned in front of the truck, and the truck's branding is visible. The scene matches the core description without contradictions.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a man wearing a black and red shirt standing next to a white Ram truck. The truck is interacting on a of the truck, and the truck is design is visible. The setting is the description description provided any.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..a224361b745d1d4a70d2365d8a8050a775b88f41
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f36238042594177e15e0f7335b1302848dd7ce04ff7157ec73f17d5ff181d4e8
+size 208865
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..d7e2142f3c26f6f9815c2f345a2e7667e3c33b17
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a soccer player in action, wearing a black and white striped jersey with the Jeep logo. The player is seen in three different positions, each showcasing a different aspect of the game. In the first frame, the player is seen pointing towards the goal, possibly strategizing with his teammates. In the second frame, the player is seen running towards the ball, indicating a high-paced game. In the third frame, the player is seen in a defensive stance, ready to intercept the ball. The background is a blur of the soccer field and the crowd, emphasizing the focus on the player's actions. The video is a dynamic representation of a soccer match, capturing the intensity and strategy involved in the game."
+    ],
+    "video_ids": [
+        "C5aVvk8Txds_59_459to668"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A soccer player in a black and white striped jersey with the Jeep logo.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a black and white striped jersey with the 'Jeep' logo clearly visible on the front. The jersey also features the Juventus crest and Adidas logo, which are consistent with the description. The player's action of pointing does not contradict the core description. The background is blurred, but this does not affect the fulfillment of the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a soccer player wearing a black and white striped jersey, the JeepJeep' logo, visible on the front. The player is features white Adidas logo, the branding, which are consistent with the description. The player is actions of kicking and not contradict the description description of The presence and a but but it does not affect the identification of the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..5c3c9049c8c73efaada888290e77405764f1d719
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1396-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3782a5de721cdbe99f28c4c8db4c59d132500b442c8f16d037bdcaf86a5d2a0
+size 410309
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..c746b617ba825deba09bd6c1560b9073d77205a2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man and a woman are engaged in a conversation at a bar. The man, wearing a black t-shirt with a graphic design, is gesturing with his hands as he speaks. The woman, dressed in a green top, is attentively listening to him. They are surrounded by an array of bottles and cans, suggesting a casual and relaxed atmosphere. The bar itself is adorned with a brick wall and framed pictures, adding to the ambiance. The scene captures a moment of interaction between the two individuals, set against the backdrop of a typical bar setting."
+    ],
+    "video_ids": [
+        "D35PWEjnmo0_77_0to103"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man and a woman? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man and a woman sitting at a table, engaging in conversation. Their presence and interaction fulfill the 'Object(s)' condition as described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a man and a woman engaged at a bar, engaging in a. The presence and interaction fulfill the 'Object(s)' condition as described.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d28899fa125ac761976f5455db51e0d4dc36ad4a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bfd07be840262e9bab948bab48d6ffc45fec3bb86e0759739d3a956bdaf44b9c
+size 173213
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..3beaf8c9fb79ee4acad654b87521a06b132ff075
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a stop-motion animation set in a Halloween-themed dollhouse. The dollhouse is purple and white with a purple door and black windows. The scene is decorated with Halloween decorations such as spider webs, pumpkins, and skeletons. The dolls are dressed in costumes and are interacting with each other and the environment. The lighting is dim with string lights adding a festive atmosphere. The style of the video is playful and whimsical, capturing the fun and creativity of Halloween celebrations."
+    ],
+    "video_ids": [
+        "T2oEHGOpKZQ_82_0to124"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Dolls dressed in costumes? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows multiple dolls dressed in various Halloween costumes, such as witches, ghosts, and other spooky outfits, which matches the 'Object(s)' condition. The costumes are clearly visible and are the central focus of the doll arrangement in the scene.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts two dolls dressed in Halloween Halloween costumes, including as a and which, and a spooky attire, which align the descriptionObject(s)' condition of The dolls are clearly visible and the the main focus of the scene figures.\"\n front video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8c007a3ad4f187fddaa3cc72f35af23a8c4089d5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:308e82f550e5d7e95bec3db82c9aa450dbcf8a1f65d60a057cbdc7c0239091ab
+size 114367
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..d0ef4dae7e9e372bb52d7653358a13cdbc2894ff
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features two animated characters, a green one and a purple one, in a playful and colorful setting. The green character is wearing a sailor's hat and has a sad expression on its face. The purple character is wearing a purple hat and has a happy expression, with its mouth wide open as if it's laughing or singing. They are standing on a sandy beach with a ship's wheel in the background, suggesting a nautical theme. The overall style of the video is cartoonish and whimsical, with bright colors and exaggerated features that give it a fun and lighthearted feel."
+    ],
+    "video_ids": [
+        "-ikqW_4eJbU_96_106to263"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two animated characters: a green one with a sailor's hat and a sad expression, and a purple one with a purple hat and a happy expression, wide open mouth.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video features two animated characters as described: a green character wearing a sailor's hat with a sad expression, and a purple character with a purple hat and a happy expression, wide open mouth. The green character's initial expression is sad, and the purple character appears cheerful and smiling. The scene also includes a lighthouse and a ship's wheel in the background, but these do not contradict the core description of the characters.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video features two animated characters as described: a green character wearing a sailor's hat and a sad expression, and a purple character wearing a purple hat and a happy expression, wide open mouth. The background character's expression expression changes indeed, but the purple character's happy. animated. The setting is includes a shipighthouse and a ship's wheel in the background, which these elements not contradict the core description of the characters.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..50e7cd958f358804ce71c07dc4204ecb458874a4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d1cb243791def14194e65646882191698bcc85af4365c9ac735ebc5173a412d6
+size 248954
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..e0b5390fddbbc9fa6938e70d9c6b2654a6020471
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a car's trunk being opened and a blue suitcase being placed inside. The trunk is spacious and appears to be well-lit, possibly by interior lights. The suitcase is large and blue, with a hard shell. The car's interior is visible, and the trunk is open, revealing the suitcase inside. The car's exterior is not visible in the video. The style of the video is straightforward and functional, focusing on the action of placing the suitcase in the trunk. The video does not contain any additional objects or people."
+    ],
+    "video_ids": [
+        "7m_0bSZJ8OQ_20_0to141"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue suitcase? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a blue suitcase placed inside the open trunk of a car, which fulfills the 'Object(s)' condition. The suitcase is the central focus and is accurately depicted as described.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a blue suitcase inside in the trunk trunk of a car. which matchesfills the 'Object(s)' condition of There presence is the primary object, there the described as blue.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..8721725681beaeda849722351bb8c449f50f3110
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50c47c475f6edc9415e8af7848ba8f5397495a2287328c49c354a95c87d28403
+size 114876
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..3bffe6936807cfd7ff051ac0ec04049c3c973c71
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases the interior of a luxury car, focusing on the driver's side. The car's interior is sleek and modern, with a black color scheme that gives it a sophisticated look. The steering wheel is prominently displayed in the first frame, with the car's logo clearly visible. In the second frame, the camera angle shifts to show the car's dashboard, which is equipped with a large touch screen display. The third frame provides a view of the car's door, which features a variety of controls and buttons. The overall style of the video is sleek and modern, with a focus on the car's interior design and features."
+    ],
+    "video_ids": [
+        "Fw81JF4XU4o_9_18to216"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Steering wheel, car's logo, dashboard, large touch screen display, car's door, various controls and buttons.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows the steering wheel with controls, the car's logo (GTR), the dashboard with a digital display, a large touch screen display, the car's door with carbon fiber accents, and various controls and buttons on the center console and steering wheel. All elements described in the condition are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video showcases shows a interior wheel, the, the car's logo,Sk),), the dashboard with a large display, a large touch screen display, the car's door with a fiber accents, and various controls and buttons on the center console and door wheel. The these match in the ' are present and visible depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..60ac11240cf01eaec2e7053d27c16fd3b9f1d6af
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:359bd6773ec7aecf12f88461d6200bf14e8d7d7443e67fe99cdf248a9e57239a
+size 222729
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..fdde40d382a8d33df126003cd9f306a19d02f820
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the vibrant beauty of a plant with yellow berries and green leaves. The plant is in full bloom, with the berries hanging from the branches, adding a pop of color to the lush greenery. The plant is set against a backdrop of a brick wall, which provides a contrasting texture to the softness of the plant. The video is a close-up shot, focusing on the plant and its details, making it a perfect study of nature's beauty. The style of the video is simple yet effective, highlighting the plant's natural beauty without any distractions."
+    ],
+    "video_ids": [
+        "h2vIUkapwY8_7_0to112"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A plant in full bloom with yellow berries and green leaves.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a plant with yellow berries and green leaves, which matches the description. The plant appears to be in full bloom with clusters of yellow berries and green foliage, including pine-like needles and other greenery. Although there are additional elements like red stems and pinecones, they do not contradict the core description and are consistent with a natural, festive plant arrangement.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a plant with yellow berries and green leaves, which align the description of The plant appears to be in full bloom, a of yellow berries, vibrant foliage. and some needles needles. broader types leaves. The there is additional elements like a bricks and a-likees, they do not contradict the core description of are part with the plant setting possibly setting setting.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ced2f7ccf9b6e9ca2a9943418cdb2eb560a10d7c
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62b192bd4cd5e66b9e4bcd6cf84844bb5578fd58d2c622a8447ef0fe2ef5ad5b
+size 256522
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..e46cafcf8b8932bed42e3008da82aa7087262d83
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is an aerial view of a city street, showcasing the urban landscape and the flow of traffic. The style of the video is a time-lapse, capturing the movement of vehicles and pedestrians over a period of time. The street is lined with buildings on both sides, and the traffic is dense, with cars and trucks moving in both directions. The buildings are of various heights and designs, reflecting the diversity of the city. The street is wide and well-maintained, with clear lanes for vehicles and sidewalks for pedestrians. The video captures the hustle and bustle of city life, with people and vehicles moving in a coordinated manner. The aerial perspective provides a unique view of the city, highlighting the layout of the streets and the density of the urban environment."
+    ],
+    "video_ids": [
+        "1VGo72NKAnw_25_0to118"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Buildings, vehicles (cars and trucks), pedestrians? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing buildings, vehicles (cars and trucks), and pedestrians. The aerial view captures multiple brick buildings along a street, cars and trucks moving or parked on the road, and pedestrians visible on sidewalks. These elements are consistent with the description provided.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting showing buildings, vehicles (cars and trucks), and pedestrians. The buildings view captures a buildings and with the street, with and trucks in along parked, the road, and pedestrians visible on the and The elements are consistent with the description provided.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..ac6f0e001169f3d277fa74b7b7b4488212ef219e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6e43d77acf067a25595bef911d4932521496710612da9c87c796e546bd33ef6b
+size 346257
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..46dfdec682487faf1884ac026c757b4daf3a8c6f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video opens with an aerial view of a serene, circular pond surrounded by lush greenery and a wooden fence. The water is crystal clear, revealing a vibrant underwater ecosystem teeming with various fish species. As the camera pans downward, it captures the dynamic movement of these fish swimming gracefully through the water. The scene transitions to an underwater perspective, showcasing a school of fish swimming in unison across the seabed. The seabed is adorned with patches of aquatic plants and rocks, adding texture and depth to the underwater environment. The fish continue their graceful dance, creating a mesmerizing display of marine life. The video concludes with a close-up shot of the fish, highlighting their intricate details and the tranquil beauty of their natural habitat."
+    ],
+    "video_ids": [
+        "3774c28e91f52bde1c7db72abf21b90eb2785e3f384c9983efe111eb310d1920"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Fish species, aquatic plants, rocks? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows various fish species swimming in the water, aquatic plants visible near the bottom, and rocks forming part of the underwater landscape. These elements align with the described 'Object(s)' condition without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts fish fish species swimming in a water, which plants floating at the bottom, and rocks at the of the underwater landscape. These elements align with the ' 'Object(s)' condition, any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..2c780698d8d21bb9dc0aea3f1f28e1a891705d04
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1397-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f2d6a8ec4ba3519279af5a457f1e78701b531539d8b023c9c44b45dbd313e01
+size 415871
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..ee2adffb2d2298f09f5afe48bdde8a5fa6c99344
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up of a plant with white flowers and green leaves. The plant is in focus and the background is blurred. The flowers are in bloom and the leaves are lush. The plant is growing in a garden and the garden is well-maintained. The plant is the main subject of the video and there are no other objects in the video. The style of the video is a close-up nature shot."
+    ],
+    "video_ids": [
+        "h_2GMByKxNQ_60_0to136"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A plant with white flowers and green leaves, which is in focus and blooming.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video prominently features a plant with white flowers and green leaves, which is clearly in focus and appears to be blooming. The flowers are detailed and vibrant, matching the description. While there is a bee and text overlay, these elements do not contradict the core description of the plant.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows features a plant with white flowers and green leaves, which are in in focus and blo to be blooming. The flowers are the and the, matching the description of The there are a slight visible additional in in these elements do not contradict the core description of the plant and}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3784fbff573ad081dcfe35cbed9748a500ecc235
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:73deba9ec5ee0cb7a1678feb70d48d72f9c9f6744204516ead9a04b1f1fc70d1
+size 127804
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..17157dfec2f44ff8e93bc2d3b320c0c702c8d50d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a silver sports car on display at a car show. The car is sleek and shiny, with a large front grille and a distinctive side vent. The car is parked on a gray carpet, and the background features a white wall with logos and text. The car is the main focus of the video, and it is presented in a way that highlights its design and features. The lighting in the video is bright, emphasizing the car's metallic finish and the reflections on its surface. The overall style of the video is sleek and modern, with a focus on the car's design and the car show setting."
+    ],
+    "video_ids": [
+        "nlEryNOR-l8_9_0to169"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A silver sports car with a large front grille and a distinctive side vent.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a silver sports car with a prominent front grille and a distinctive side vent, matching the description. The car's design features, including the grille and side vent, are clearly visible and align with the specified characteristics.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows a silver sports car with a sleek front grille and a distinctive side vent, which the description provided The car's sleek features, such the grille and side vent, are clearly visible and align with the given characteristics.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..72e94927200152f03fdf42798265ed70edde50ae
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ffb0bc09336112bdaade632df6f2753f2bc0decf62c2a417cd681bbff1e66c29
+size 228829
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..dc1d2a27d7b917d14ab6f07636d2ebb2a5f5b1f2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a woman with red hair, wearing a black jacket and a purple shirt. She is standing in front of a red bulletin board with yellow borders. The woman appears to be speaking, as her mouth is open and her expression is one of concern or seriousness. The background is blurred, but it seems to be an indoor setting, possibly a classroom or office. The style of the video is candid and informal, capturing a moment in the woman's day."
+    ],
+    "video_ids": [
+        "OTHB7ubXbGs_41_0to120"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman with red hair, wearing a black jacket and a purple shirt; a red bulletin board with yellow borders.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with red hair, wearing a black jacket and a purple shirt, which matches the description. Behind her, there is a red bulletin board with yellow borders, also matching the description. No elements contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a woman with red hair wearing wearing a black jacket and a purple shirt. standing matches the description. Additionally her, there is a red bulletin board with yellow borders, also matching the description. The additional contradict the given description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..90c1cd960a16b8f18d90729ac45276a8cf2b6614
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:779d90a3d2f3f54acae07d8f7338ca3fae85c9226938e91b55004dde1d1834a1
+size 165235
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..daba67442d839cd8cea2e3ab2ede6b87b2efd573
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a playful and colorful animation featuring a young girl and two toy cars. The girl, dressed in a vibrant red outfit, stands in the center of the frame, holding out her hand as if inviting the cars to join her. To her right, a blue car with a friendly face is parked, while to her left, a yellow car with a similar face is also present. The background is a lively green, filled with trees and bushes, creating a sense of a lush, outdoor setting. The overall style of the video is whimsical and child-friendly, with bright colors and simple, cartoon-like characters."
+    ],
+    "video_ids": [
+        "LcgHRIIbme8_21_237to360"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Young girl, blue car with a friendly face, yellow car with a similar face? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a young girl figurine, a blue car with a friendly face, and a yellow car with a similar face, all of which match the described 'Object(s)' condition. The background and additional elements do not contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a young girl inine in a blue car with a friendly face, and a yellow car with a similar face. all of which match the description objectsObject(s)' condition. The presence and setting elements, not contradict the description description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..298b7f65c3ca982ceab2b438a144977cb306651e
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eaed6ee8aac16b4ba1907ab444bd6eaf75fdf4d2197c8dced27d01f97e8294b4
+size 96542
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..601cbd12217157e5b4d0208fd110944392eb0ba2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a car in a garage, with a focus on its front end. The car is a sports car with a sleek design, featuring a combination of orange and black colors. The car is parked in a garage with other cars in the background. The garage has a concrete floor and metal beams, giving it an industrial look. The lighting in the garage is bright, highlighting the car's design and color scheme. The car appears to be in good condition, suggesting it is well-maintained. The video style is straightforward, with no additional elements or effects, focusing solely on the car and its surroundings."
+    ],
+    "video_ids": [
+        "DI87TtB3QPc_49_85to291"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A sports car with a sleek design, featuring a combination of orange and black colors, parked in the garage.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a sports car with a sleek design, prominently featuring a combination of orange and black colors. It is parked inside a garage, surrounded by other vehicles and garage equipment, which aligns with the description. The car's distinctive front and rear styling, including the orange accents and black body, matches the specified color scheme and design.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a sports car with a sleek design, predominantly featuring a combination of orange and black colors. The is parked in a garage, which by other vehicles, a equipment, which iss with the description of The car's design color grille side styling, along its headlights and and black body, are the given colors combination. design elements}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..960b7126dc30eaf7932581319b1cb718348bba9f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a12a284dac9d5a19a0fa4484926f0b74e0b9d6ddec7f23919473856e2d34251f
+size 67475
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..01355bf9327e4c9109dd7ba375ae1a0423cc7eef
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a close-up view of a silver wheel with a black tire, which is part of a vehicle. The wheel is positioned in the center of the frame, and the tire appears to be new and well-maintained. The background features a brick wall with a sign that reads \"M&M Motors\". The sign is yellow and black, and it is mounted on the wall. The style of the video is a straightforward, close-up shot of the wheel and tire, with no additional action or movement. The focus is solely on the wheel and tire, and the background is kept simple and uncluttered. The lighting in the video is bright and even, highlighting the details of the wheel and tire. The overall impression is that of a professional, high-quality video showcasing the wheel and tire."
+    ],
+    "video_ids": [
+        "KeAKp-d2ZFc_5_0to186"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A silver wheel with a black tire, a brick wall, and a yellow and black sign.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a silver wheel with a black tire as the central focus, a brick wall in the background, and a yellow and black sign (Magnaflo logo) visible on the wall. These core elements are accurately represented, fulfilling the 'Object(s)' condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a silver wheel with a black tire, the main focus. which brick wall in the background, and a yellow and black sign partiallypartoliay)) partially above the wall. The elements elements match present represented in fulfilling the 'Object(s)' condition.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..47663597e7bb73f542793be3efaf9b1abac8e7ab
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ace2c5636019827cd8bf9f238385a45285113d0b393869ff34b21ede6bf1dc95
+size 97552
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..8cd03914c1206cfd25594686ace4418a456840b4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is an aerial view of a small town, showcasing its layout and architecture. The town is characterized by a grid of streets lined with houses and buildings, with a central road that runs through the middle. The buildings are predominantly one-story structures, with a mix of residential and commercial properties. The town is surrounded by a lush green landscape, with trees and fields visible in the background. The video captures the town's tranquility and charm, with no visible movement or activity. The style of the video is realistic, with a focus on the town's layout and architecture. The aerial perspective provides a comprehensive view of the town, allowing viewers to appreciate its layout and design. The video is likely intended for use in real estate or tourism promotion, highlighting the town's unique features and charm."
+    ],
+    "video_ids": [
+        "dNU1lJiDaSY_3_23to203"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Houses, buildings, central road, trees, fields? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing houses, various buildings, a central road with vehicles, trees scattered throughout the area, and fields in the background. All these elements are visible and consistent with the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as depicting depicting houses, buildings buildings, a central road, a, trees, throughout the area, and fields in the background. The these elements are present and contribute with the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..9810df7bb2a9d16d024a9efca2fd9e1c6a0f6845
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f0f85ffd3f89856b88da1d74ac8d2ad99fe793e3d42ca5089c6426828df9feca
+size 278830
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..a71a4fa9916b362c3510777a461d1f3334064d00
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two women are cooking in a kitchen. The woman on the left is stirring something in a pot, while the woman on the right is holding her head, possibly in frustration or disbelief. The kitchen is well-equipped with various appliances and utensils, including a refrigerator, an oven, a sink, and a cutting board. There are also several bowls and bottles on the counter. The scene suggests that the women are in the middle of preparing a meal, and the woman on the right might be experiencing some difficulty or confusion in the process. The overall style of the video is realistic and candid, capturing a moment of everyday life."
+    ],
+    "video_ids": [
+        "NH0P-u-8dzc_34_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two women, one stirring something in a pot and the other holding her head.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two women in a kitchen setting. One woman is actively stirring something in a pot on the stove, while the other woman has her hands on her head, appearing distressed or overwhelmed. This matches the described condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two women in a kitchen setting. One woman is actively stirring something in a pot, the stove, which the other woman is her hand on her head, appearing to or overwhelmed. The matches the description scenario of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..c872625a6705f3bc4f120448a741989f81a9bbd4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1398-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88b98f2381acf2ca72de356996d02c4f5621597596194f230f7e5c8bcff726fb
+size 119340
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..de4dbce9e9235a68e70bc7dec83d7765865a3aed
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a close-up of a piece of meat being cooked on a grill. The meat is browned and has a crusty exterior, indicating it is being cooked at a high temperature. The grill has a temperature gauge, which is set to 350 degrees, suggesting that the meat is being cooked at a high heat. The grill is made of metal and has a black handle. The meat is placed on a metal grate, which is placed on the grill. The grill is placed on a wooden table. The table is brown and has a grainy texture. The background is blurred, but it appears to be an outdoor setting. The style of the video is realistic and it captures the details of the cooking process."
+    ],
+    "video_ids": [
+        "glt8Jn2MKgU_18_35to158"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Grill, meat, temperature gauge, metal grate? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a grill with meat on a metal grate, and a temperature gauge (specifically a Kamado Joe thermometer) is prominently displayed. All specified objects are present and accurately depicted without contradiction.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows fulfills the 'Object(s)' condition as showing showing a grill, a on it metal grate. and a temperature gauge istherally a meatado Joe thermometer) is visible displayed, The the objects are present and correctly depicted in any.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..0f8c8eb9f745cff5b904bfc3e38c29f5e5062fbc
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07fb0bf2c53426b5691f3ced1dc9c118d323d814a9da3d92ca9e7ffda4748919
+size 157474
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..5b274b0f9b7c03cad0cdf8c0edddb3216a2313a0
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a dynamic moment in a basketball game between the Oklahoma City Thunder and the Houston Rockets. The main focus is on two players, one from each team, engaged in a fierce competition. The player from the Oklahoma City Thunder, dressed in an orange jersey, is in a defensive stance, his eyes locked onto the ball. His posture suggests he's ready to intercept or block the ball. On the other hand, the player from the Houston Rockets, wearing a black jersey, is holding the ball, preparing to make a move. His body language indicates he's strategizing his next move. The background is a blur of spectators, indicating the game is being played in a packed stadium. The atmosphere is intense, reflecting the high stakes of the game."
+    ],
+    "video_ids": [
+        "JyTle8ygHio_24_24to175"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two players, one from Oklahoma City Thunder (orange jersey) and one from Houston Rockets (black jersey).? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two basketball players in a game setting. One player is wearing an orange jersey with \\\"OKC 21\\\" which corresponds to the Oklahoma City Thunder, and the other is wearing a black jersey with \\\"Houston\\\" visible, which corresponds to the Houston Rockets. The core description of the two players from these teams is accurately represented.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows two basketball players, action game setting. One player is wearing an orange jersey with theOKC\\\"6\\\"\\\" on is to the Oklahoma City Thunder, and the other player wearing a black jersey with \\\"K Rockets written, which corresponds to the Houston Rockets. The players description of the ' players and the teams is fulfilled represented in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..3bd98574e61179fff3beb3611f6d7d70af2c5dae
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2ac1b4cf7d61ffa8f801cfcfa0155dd2758ef73fb5275c29c53794129b48b79a
+size 325545
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..5b819bfc07ec023798605e6669c7ee8f4c41c562
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a man is seen sitting in the driver's seat of a car. He is wearing a denim jacket and a red lanyard. The car has a red interior, and the man is gesturing with his hand towards the camera. The car is parked in a garage, and there are stairs visible in the background. The man appears to be speaking, and his expression suggests that he is engaged in a conversation. The overall style of the video is casual and informal, with a focus on the man and his interaction with the camera."
+    ],
+    "video_ids": [
+        "0XL0HWZ1Zg0_69_114to250"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a denim jacket with a red lanyard, a car with a red interior? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a denim jacket and a red lanyard, seated inside a car with a red interior, specifically a red racing-style seat. These elements align with the description, even though the car's interior may not be fully visible, the red seat is clearly identifiable. The presence of additional background elements does not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man wearing a denim jacket and a red lanyard, sitting inside a car with a red interior. which the red seat-style seat. The elements match with the description provided making though the car's interior is not be fully visible, the red seat is a identifiable. The presence of the elements elements in not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..49793bd87ccd0bddffaa8544b978e6f5cc51d2c5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7009436bdde1ae70eca52c46506d266a2bab8106947f582547581db2f6b967bb
+size 212329
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..3b4be1f4047fee4362f61acee4a9c2d0725accf7
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, two young boys are seen opening Christmas presents in a living room. The first boy, wearing a blue shirt, is seen holding a present and looking at it with curiosity. The second boy, wearing a black shirt, is seen sitting on the floor, watching the first boy with interest. The living room is filled with various Christmas presents, some of which are wrapped in blue and white wrapping paper. The boys are surrounded by family members who are watching them with smiles on their faces. The atmosphere is festive and joyful, with the excitement of Christmas present-giving."
+    ],
+    "video_ids": [
+        "C3_MLqOCACM_18_207to379"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Two young boys and family members? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows two young boys sitting on the floor unwrapping a gift, with several adults (family members) seated on couches in the background. The core description of 'Two young boys and family members' is accurately represented, with no conflicting elements.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts two young boys in on the floor surroundedrapping gifts gift, with two wrapped inlikely members) seated on aes in the background. The scene elements of 'Two young boys and family members' is fulfilled represented in with the significant elements.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..af9802a1776017733d553d8003af53d91d0bdba5
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:136f089ddd1ce4bd81a66ac7c003600085d234f8d129def4334a3c4c10a9d7e3
+size 195465
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..15cde249fb996dc11f556f63d75bd6cc08e18200
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a television show featuring a woman in the foreground, who is the main subject. She is wearing a blue sweater and a scarf, and she is speaking. In the background, there is a studio audience, which includes several people sitting in chairs. The audience members are dressed in various styles of clothing, and they appear to be listening to the woman. The style of the video is a standard television show format, with a focus on the woman in the foreground and the audience in the background. The lighting in the studio is bright, and the overall atmosphere is one of a live television broadcast."
+    ],
+    "video_ids": [
+        "CfiU_PtbZfI_1_0to165"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A woman in the foreground wearing a blue sweater and scarf, and a studio audience in the background consisting of several seated individuals dressed in various styles of clothing.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a woman in the foreground wearing a blue sweater and scarf, and a studio audience in the background with several seated individuals dressed in various styles of clothing. These elements align with the described 'Object(s)' condition, even though additional graphical overlays (like the Steve Harvey promotion) are present, which do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows depicts a woman in the foreground wearing a blue sweater and a. which a studio audience in the background consisting several seated individuals dressed in various styles of clothing. The elements match with the description 'Object(s)' condition without making though there elements elements oflike the blue Jobs image) are present, they do not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..08fc25f216dcb1afe596f83bb8040f9920cf2c6a
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:499469bfb6bc88f4f163c8a0c6afcb342bc811d8b6dba92f73e40fcacc854962
+size 169010
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..58804b7145e0d8efa3377c173f39ea23c7516ac1
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures the dynamic movement of a blue sports car as it drives down a road. The car's sleek design and shiny exterior reflect the sunlight, creating a sense of speed and motion. The car's large black wheels with silver spokes add to its sporty aesthetic. The road beneath the car is a blur, indicating its high speed. The background is a blur of trees and foliage, suggesting that the car is driving through a forest or a park. The overall style of the video is dynamic and energetic, capturing the thrill of driving a sports car."
+    ],
+    "video_ids": [
+        "j8yfg2DpLHQ_56_0to144"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A blue sports car? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a blue sports car, specifically an Aston Martin, with detailed shots of its front wheel, grille, headlights, and bodywork. The car's color and sporty design match the description, and there are no elements that contradict this core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a blue sports car in which focusing electric Martin, as a features of its side wheel and body, and, and part.. The car's design and designy design align the description of and there are no additional that contradict the description description.\"\n}<|im_end|>\n<|im_start|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-5.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-5.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..d66e76b3e8abe061bb6b9e32c722f6d0273dc438
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-5.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e1dbfa5478866e7c6139b95fcac0d6f825cea58033c02119c355e521366074d
+size 269706
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-6.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-6.json
new file mode 100644
index 0000000000000000000000000000000000000000..713ca8bb816fad4a7e8615b6f6da968c16230915
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-6.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video captures a close-up view of a vibrant salad being prepared in a wooden bowl. The salad is a colorful mix of fresh ingredients, including lettuce, tomatoes, carrots, and onions. The ingredients are chopped and mixed together, creating a visually appealing dish. The wooden bowl adds a rustic touch to the scene, and the salad is garnished with a sprig of parsley. The video is shot in a realistic style, focusing on the textures and colors of the ingredients. The salad is the main subject of the video, and the camera angle is close enough to capture the details of the ingredients. The video does not contain any text or additional elements. The overall style of the video is simple and straightforward, focusing on the preparation of the salad."
+    ],
+    "video_ids": [
+        "qsdEl4Uj6jg_23_0to164"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Wooden bowl, lettuce, tomatoes, carrots, onions, parsley? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a wooden bowl containing a salad with lettuce, cherry tomatoes, orange carrot pieces, sliced red onions, and green parsley. All the specified elements are present and accurately represented in the visual content.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows shows a wooden bowl containing a salad with lettuce, tomatoes tomatoes, sliced slices slices, chopped onions onions, and parsley parsley. The the elements objects are present, match depicted in the image content.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-6.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-6.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7e31406b9e4b3e2e2153fe209471d8f5669cf493
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-6.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac375770868a7a275d64b43353d0c20e1126393fa2aaf84a00aab9082af96626
+size 139252
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-7.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-7.json
new file mode 100644
index 0000000000000000000000000000000000000000..b98112c09d571c87fe40871c808e5dae641886af
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-7.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video features a man in a suit and tie, sitting in a chair and gesturing with his right hand. He appears to be speaking or making a point. The background is a blue wall with the word \"ABRA\" written on it. The man is wearing a blue suit and tie, and his hair is gray. The style of the video is a news or interview segment, with the man likely being a guest or commentator. The focus is on the man and his expression and gestures, with the background serving as a simple backdrop."
+    ],
+    "video_ids": [
+        "4pnHU8MksA0_11_121to265"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man in a suit and tie, gesturing with his right hand.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man dressed in a suit and tie, and he is actively gesturing with his right hand, pointing his index finger. This matches the described condition.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a man in in a suit and tie, and he is indeed gesturing with his right hand. which and finger finger. This matches the description ' of}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-7.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-7.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b027f8aa5f2289ab2e19a90676162c234102a24d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-1399-7.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:93096f49f38bbab2891415c0b35ad62a6e0cdf40a5dc1fbff8e750afe4f4b38e
+size 172668
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-0.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-0.json
new file mode 100644
index 0000000000000000000000000000000000000000..35f392e62258788c31cee0105f3f7ea5d8b97fb4
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-0.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a man enjoying a meal at a restaurant. He is wearing glasses and a striped shirt. In the first frame, he is holding a spoonful of food to his mouth. In the second frame, he is taking a bite of the food. In the third frame, he is looking at the camera with a smile on his face. The food on his plate appears to be a bowl of rice with some vegetables. The restaurant has a floral wallpaper on the wall behind him. The overall style of the video is casual and relaxed, capturing a simple yet enjoyable moment of dining."
+    ],
+    "video_ids": [
+        "BozPRwYOJrw_28_0to145"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A man, a spoon, a bowl of rice, some vegetables, the man's glasses, and a striped shirt.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a man wearing glasses and a striped shirt, eating from a bowl of rice with a spoon. The bowl contains rice and some visible vegetables (like green onions and possibly shrimp). All specified objects are present and accurately depicted in the video.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"No\",\n  \"explanation\": \"The video shows depicts a man eating glasses and a striped shirt, holding from a spoon with food with some spoon. The bowl contains rice and some vegetables vegetables.yellow carrots and and possibly carrots). The the elements are present and match depicted in the video.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-0.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-0.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..723ab01af474c4362048ab59c9fe2603c3af4c73
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-0.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8a2c76ba762901ffd2eedf983b4a7c1c5b70fe88d11473142250faab358c63ce
+size 317907
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-1.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-1.json
new file mode 100644
index 0000000000000000000000000000000000000000..d8e08d484d8cd95c472d8b89c681ff49c9718713
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-1.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a single fish swimming gracefully through the deep blue ocean. The fish, likely a triggerfish given its distinctive features, has a robust body covered in a pattern of dark and light scales. Its fins are large and fanned out, with the pectoral fins displaying a striking yellow and black coloration. The fish's tail fin is also prominently visible, with a similar pattern of yellow and black stripes. The background is a uniform deep blue, indicating the depth of the water and the absence of any other marine life or objects in the immediate vicinity. The fish moves smoothly from left to right across the frame, its body undulating gently as it navigates through the water. The BBC logo is visible in the top left corner, suggesting that this footage is part of a documentary or educational program. The overall scene is serene and focused on the solitary fish, highlighting its natural beauty and the tranquil underwater environment."
+    ],
+    "video_ids": [
+        "717f194ea931acb6bda1527dbe8f4d790e00e6f9e58506ea89646ddc7d95eeba"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A single triggerfish? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a single fish swimming in the ocean, which matches the description of a triggerfish. The fish has the characteristic body shape, coloration, and fin structure of a triggerfish, and there are no conflicting elements in the video that contradict this description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a single trigger that in the water, which align the description of a 'fish. The fish is the characteristic features shape and coloration, and fin structure of a triggerfish. and there are no other elements in the video.\"\n would the description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-1.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-1.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..b22f6ae072eb6a48341e02f6c3c03ac42ff28caf
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-1.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1244a753e3ed5db139fbc849f2a93d3d14700a9944cbbf1e354b44fb00fe459e
+size 248442
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-2.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-2.json
new file mode 100644
index 0000000000000000000000000000000000000000..4722d1790c9d1227917b0454e4129911d24a04a2
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-2.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "In the video, a dog and an owl are seen interacting with each other amidst a pile of autumn leaves. The dog, with its black fur, is seen peeking out from the leaves, while the owl, with its distinctive brown and white feathers, is perched on the leaves. The leaves are a mix of orange and yellow, indicating the season is fall. The dog and owl seem to be curious about each other, with the dog looking at the owl and the owl looking back. The overall scene is peaceful and serene, capturing the beauty of nature and the interaction between two different species."
+    ],
+    "video_ids": [
+        "A5PPGTdnpPU_27_0to199"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: A dog and an owl? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video clearly shows a dog and an owl partially buried in autumn leaves, matching the core description. The presence of leaves is acceptable as it fits the natural setting and does not contradict the main subjects.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts depicts a dog and an owl, visible in autumn leaves, which the ' description of The presence of the and an as it adds the autumn setting and does not contradict the main objects,}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-2.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-2.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..91d120cc8fd9153923bf7e3c0b7148589f332401
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-2.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5ed12a68fc4123942451e2f850c007528414b2d3b07c0957f50e0d820979efc
+size 181317
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-3.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-3.json
new file mode 100644
index 0000000000000000000000000000000000000000..470bf3d7c2a04c25519fd0b7a2d1941bcb72f73f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-3.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video shows a cozy bedroom scene with a large bed and a window view of a tropical garden. The bed is neatly made with a green comforter and matching pillows. A wooden nightstand with a lamp is placed next to the bed. A framed picture hangs above the bed, adding a touch of personality to the room. The room has a warm and inviting atmosphere, with the natural light from the window creating a serene ambiance. The tropical garden outside the window adds a touch of nature to the room, making it feel more like a retreat. The overall style of the video is calm and relaxing, perfect for a peaceful evening in."
+    ],
+    "video_ids": [
+        "YP9q1b5LHTs_66_0to145"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Large bed, green comforter, matching pillows, wooden nightstand, lamp, framed picture, tropical garden view through the window? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully fulfills the 'Object(s)' condition by clearly showing a large bed with a green comforter, matching patterned pillows, wooden nightstands on either side, a lamp on one nightstand, a framed picture above the bed, and a view of a tropical garden through the windows. All specified elements are present and accurately depicted.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts fulfills the 'Object(s)' condition as showcasing showing a large bed with a green comforter, matching pillowsed pillows, a nightstand, either side, a lamp on one ofstand, a framed picture on the bed, and a tropical of a tropical garden through the window. The elements elements are present and match depicted in}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-3.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-3.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..7cd40050611f3b6b24329dc4fe3a5e5ab576916f
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-3.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8dc841d21cbb0a1ae38feb5f9a481118ae8142f34d484496678219dc466387de
+size 238876
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-4.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-4.json
new file mode 100644
index 0000000000000000000000000000000000000000..34a72b638565c5444f0d08ed34abd10705373a4d
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-4.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video showcases a Formula 1 car in a studio setting. The car is predominantly orange and black, with the number 7 on the side. It features a blue and white front wing and a blue and white rear wing. The car is equipped with a black helmet and a black steering wheel. The car is stationary, and there is no text or other objects in the video. The style of the video is a realistic 3D rendering, providing a detailed and accurate representation of the Formula 1 car."
+    ],
+    "video_ids": [
+        "AEr2Kdj5Vsc_12_0to114"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: Formula 1 car, black helmet, black steering wheel? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows a Formula 1 car, which is the primary subject. The driver is wearing a black helmet, and the steering wheel is also black, matching the description. Additional elements like sponsor logos and the car's livery do not contradict the core description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video depicts a Formula 1 car with which is the primary object. The car is wearing a black helmet, and the steering wheel is not black. as the description. The elements like the logos and the car's designivery are not contradict the core description.\"\n}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|im_start|>"
+    ]
+}
\ No newline at end of file
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-4.mp4 b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-4.mp4
new file mode 100644
index 0000000000000000000000000000000000000000..429101a553237ae31134c8b832693965a5c54504
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-4.mp4
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed46232712e05f70e6002c1ecaac159f8d3fb42b5a6e6e6d228ccdc8608a7143
+size 103541
diff --git a/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-5.json b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-5.json
new file mode 100644
index 0000000000000000000000000000000000000000..7785b240c898951bb089972ea8cedb163054bbb3
--- /dev/null
+++ b/VideoX-Fun/output_sd_objects/train_sample_full/_hf_sharded/shard_0001/sample-14-5.json
@@ -0,0 +1,18 @@
+{
+    "prompt": [
+        "The video is a promotional advertisement for boating, showcasing the interior of a luxury boat. The boat's dashboard is prominently displayed, featuring a variety of gauges and controls. The steering wheel is large and centrally located, with a polished finish. The boat's interior is spacious and well-lit, with comfortable seating and a luxurious feel. The boat is on the water, with a serene view of the lake and surrounding trees. The video is shot in a realistic style, with high-quality photography and a focus on the boat's features and the surrounding environment."
+    ],
+    "video_ids": [
+        "XYDY2FdSt9g_8_0to115"
+    ],
+    "question": [
+        "Given this AI-generated video, does it successfully fulfill the 'Object(s)' condition: The boat's dashboard, featuring gauges and controls; the steering wheel, which is large and centrally located with a polished finish; comfortable seating; and the lake and trees outside the boat.? Respond with 'Yes' or 'No', Answer 'Yes' if the video largely matches the description. Answer 'No' if the video clearly contradicts the description. The presence of additional elements in the video is acceptable as long as they do not conflict with the core description. Please provide a brief explanation for your answer. \nProvide your analysis and explanation in JSON with keys: answer (e.g., Yes or No), explanation."
+    ],
+    "gt_answer": [
+        "{\n  \"answer\": \"Yes\",\n  \"explanation\": \"The video successfully shows the boat's dashboard with gauges and controls, a large centrally located steering wheel with a polished finish, comfortable seating, and the lake and trees visible outside the boat. All described elements are clearly present and match the description.\"\n}"
+    ],
+    "pred_answer": [
+        "  \"answer\": \"Yes\",\n  \"explanation\": \"The video shows ful the boat's dashboard with variousuges and controls, the large and located steering wheel with a polished finish, comfortable seating, and a lake and trees outside outside the boat. The elements elements are present present in match the description provided}<|im_end|>\n<|endoftext|>",
+        "  \"has_obvious_defect\"\": No,\n  \"\"dominant_issue\"\": \"\"none\"\",\n detail_loss_vs_reference\"\": NoNo\"\",\n evidence\"\": [}<|im_end|>\n<|endoftext|>"
+    ]
+}
\ No newline at end of file